id	sid	tid	token	lemma	pos
dpc-1863	1	1	dermatology	dermatology	NOUN
dpc-1863	1	2	:	:	PUNCT
dpc-1863	1	3	practical	practical	ADJ
dpc-1863	1	4	and	and	CCONJ
dpc-1863	1	5	conceptual	conceptual	ADJ
dpc-1863	1	6	original	original	ADJ
dpc-1863	1	7	article	article	NOUN
dpc-1863	1	8	|	|	ADV
dpc-1863	1	9	dermatol	dermatol	ADJ
dpc-1863	1	10	pract	pract	NOUN
dpc-1863	1	11	concept	concept	NOUN
dpc-1863	1	12	.	.	PUNCT
dpc-1863	2	1	2022;12(3):e2022126	2022;12(3):e2022126	NUM
dpc-1863	2	2	1	1	NUM
dpc-1863	2	3	dermatology	dermatology	NOUN
dpc-1863	2	4	practical	practical	ADJ
dpc-1863	2	5	&	&	CCONJ
dpc-1863	2	6	conceptual	conceptual	ADJ
dpc-1863	2	7	comparison	comparison	NOUN
dpc-1863	2	8	of	of	ADP
dpc-1863	2	9	convolutional	convolutional	ADJ
dpc-1863	2	10	neural	neural	ADJ
dpc-1863	2	11	network	network	NOUN
dpc-1863	2	12	architectures	architecture	NOUN
dpc-1863	2	13	for	for	ADP
dpc-1863	2	14	robustness	robustness	NOUN
dpc-1863	2	15	against	against	ADP
dpc-1863	2	16	common	common	ADJ
dpc-1863	2	17	artefacts	artefact	NOUN
dpc-1863	2	18	in	in	ADP
dpc-1863	2	19	dermatoscopic	dermatoscopic	NOUN
dpc-1863	2	20	images	image	NOUN
dpc-1863	2	21	florian	florian	ADJ
dpc-1863	2	22	katsch1	katsch1	PROPN
dpc-1863	2	23	,	,	PUNCT
dpc-1863	2	24	christoph	christoph	PROPN
dpc-1863	2	25	rinner1	rinner1	PROPN
dpc-1863	2	26	,	,	PUNCT
dpc-1863	2	27	philipp	philipp	PROPN
dpc-1863	2	28	tschandl2	tschandl2	PROPN
dpc-1863	2	29	1	1	NUM
dpc-1863	2	30	 	 	SPACE
dpc-1863	2	31	center	center	NOUN
dpc-1863	2	32	for	for	ADP
dpc-1863	2	33	medical	medical	ADJ
dpc-1863	2	34	statistics	statistic	NOUN
dpc-1863	2	35	,	,	PUNCT
dpc-1863	2	36	informatics	informatic	NOUN
dpc-1863	2	37	and	and	CCONJ
dpc-1863	2	38	intelligent	intelligent	ADJ
dpc-1863	2	39	systems	system	NOUN
dpc-1863	2	40	,	,	PUNCT
dpc-1863	2	41	medical	medical	ADJ
dpc-1863	2	42	university	university	PROPN
dpc-1863	2	43	of	of	ADP
dpc-1863	2	44	vienna	vienna	PROPN
dpc-1863	2	45	,	,	PUNCT
dpc-1863	2	46	vienna	vienna	PROPN
dpc-1863	2	47	,	,	PUNCT
dpc-1863	2	48	austria	austria	PROPN
dpc-1863	2	49	2	2	NUM
dpc-1863	2	50	 	 	SPACE
dpc-1863	2	51	department	department	NOUN
dpc-1863	2	52	of	of	ADP
dpc-1863	2	53	dermatology	dermatology	NOUN
dpc-1863	2	54	,	,	PUNCT
dpc-1863	2	55	medical	medical	ADJ
dpc-1863	2	56	university	university	PROPN
dpc-1863	2	57	of	of	ADP
dpc-1863	2	58	vienna	vienna	PROPN
dpc-1863	2	59	,	,	PUNCT
dpc-1863	2	60	vienna	vienna	PROPN
dpc-1863	2	61	,	,	PUNCT
dpc-1863	2	62	austria	austria	PROPN
dpc-1863	2	63	key	key	ADJ
dpc-1863	2	64	words	word	NOUN
dpc-1863	2	65	:	:	PUNCT
dpc-1863	2	66	image	image	NOUN
dpc-1863	2	67	classification	classification	NOUN
dpc-1863	2	68	,	,	PUNCT
dpc-1863	2	69	object	object	NOUN
dpc-1863	2	70	detection	detection	NOUN
dpc-1863	2	71	,	,	PUNCT
dpc-1863	2	72	instance	instance	NOUN
dpc-1863	2	73	segmentation	segmentation	NOUN
dpc-1863	2	74	,	,	PUNCT
dpc-1863	2	75	artefacts	artefact	NOUN
dpc-1863	2	76	,	,	PUNCT
dpc-1863	2	77	dermatoscopy	dermatoscopy	NOUN
dpc-1863	2	78	,	,	PUNCT
dpc-1863	2	79	citation	citation	NOUN
dpc-1863	2	80	:	:	PUNCT
dpc-1863	2	81	katsch	katsch	PROPN
dpc-1863	2	82	f	f	X
dpc-1863	2	83	,	,	PUNCT
dpc-1863	2	84	rinner	rinner	NOUN
dpc-1863	2	85	c	c	NOUN
dpc-1863	2	86	,	,	PUNCT
dpc-1863	2	87	tschandl	tschandl	NOUN
dpc-1863	3	1	p.	p.	NOUN
dpc-1863	3	2	comparison	comparison	NOUN
dpc-1863	3	3	of	of	ADP
dpc-1863	3	4	convolutional	convolutional	ADJ
dpc-1863	3	5	neural	neural	ADJ
dpc-1863	3	6	network	network	NOUN
dpc-1863	3	7	architectures	architecture	NOUN
dpc-1863	3	8	for	for	ADP
dpc-1863	3	9	robustness	robustness	NOUN
dpc-1863	3	10	against	against	ADP
dpc-1863	3	11	common	common	ADJ
dpc-1863	3	12	artefacts	artefact	NOUN
dpc-1863	3	13	in	in	ADP
dpc-1863	3	14	dermatoscopic	dermatoscopic	ADJ
dpc-1863	3	15	images	image	NOUN
dpc-1863	3	16	.	.	PUNCT
dpc-1863	4	1	dermatol	dermatol	NOUN
dpc-1863	4	2	pract	pract	PROPN
dpc-1863	4	3	concept	concept	NOUN
dpc-1863	4	4	.	.	PUNCT
dpc-1863	5	1	2022;12(3):e2022126	2022;12(3):e2022126	X
dpc-1863	5	2	.	.	PUNCT
dpc-1863	5	3	doi	doi	NOUN
dpc-1863	5	4	:	:	PUNCT
dpc-1863	6	1	https://doi.org/10.5826/dpc.1203a126	https://doi.org/10.5826/dpc.1203a126	NOUN
dpc-1863	6	2	accepted	accept	VERB
dpc-1863	6	3	:	:	PUNCT
dpc-1863	6	4	december	december	PROPN
dpc-1863	6	5	7	7	NUM
dpc-1863	6	6	,	,	PUNCT
dpc-1863	6	7	2021	2021	NUM
dpc-1863	6	8	;	;	PUNCT
dpc-1863	6	9	published	publish	VERB
dpc-1863	6	10	:	:	PUNCT
dpc-1863	7	1	july	july	PROPN
dpc-1863	7	2	2022	2022	NUM
dpc-1863	7	3	copyright	copyright	NOUN
dpc-1863	7	4	:	:	PUNCT
dpc-1863	7	5	©	©	ADJ
dpc-1863	7	6	2022	2022	NUM
dpc-1863	7	7	katsch	katsch	NOUN
dpc-1863	7	8	et	et	PROPN
dpc-1863	7	9	al	al	PROPN
dpc-1863	7	10	.	.	PUNCT
dpc-1863	8	1	this	this	PRON
dpc-1863	8	2	is	be	AUX
dpc-1863	8	3	an	an	DET
dpc-1863	8	4	open	open	ADJ
dpc-1863	8	5	-	-	PUNCT
dpc-1863	8	6	access	access	NOUN
dpc-1863	8	7	article	article	NOUN
dpc-1863	8	8	distributed	distribute	VERB
dpc-1863	8	9	under	under	ADP
dpc-1863	8	10	the	the	DET
dpc-1863	8	11	terms	term	NOUN
dpc-1863	8	12	of	of	ADP
dpc-1863	8	13	the	the	DET
dpc-1863	8	14	creative	creative	ADJ
dpc-1863	8	15	commons	common	NOUN
dpc-1863	8	16	attributionnoncommercial	attributionnoncommercial	ADJ
dpc-1863	8	17	license	license	NOUN
dpc-1863	8	18	(	(	PUNCT
dpc-1863	8	19	by	by	ADP
dpc-1863	8	20	-	-	PUNCT
dpc-1863	8	21	nc-4.0	nc-4.0	NOUN
dpc-1863	8	22	)	)	PUNCT
dpc-1863	8	23	,	,	PUNCT
dpc-1863	8	24	https://creativecommons.org/licenses/by-nc/4.0/	https://creativecommons.org/licenses/by-nc/4.0/	PROPN
dpc-1863	8	25	,	,	PUNCT
dpc-1863	8	26	which	which	PRON
dpc-1863	8	27	permits	permit	VERB
dpc-1863	8	28	unrestricted	unrestricted	ADJ
dpc-1863	8	29	noncommercial	noncommercial	ADJ
dpc-1863	8	30	use	use	NOUN
dpc-1863	8	31	,	,	PUNCT
dpc-1863	8	32	distribution	distribution	NOUN
dpc-1863	8	33	,	,	PUNCT
dpc-1863	8	34	and	and	CCONJ
dpc-1863	8	35	reproduction	reproduction	NOUN
dpc-1863	8	36	in	in	ADP
dpc-1863	8	37	any	any	DET
dpc-1863	8	38	medium	medium	NOUN
dpc-1863	8	39	,	,	PUNCT
dpc-1863	8	40	provided	provide	VERB
dpc-1863	8	41	the	the	DET
dpc-1863	8	42	original	original	ADJ
dpc-1863	8	43	authors	author	NOUN
dpc-1863	8	44	and	and	CCONJ
dpc-1863	8	45	source	source	NOUN
dpc-1863	8	46	are	be	AUX
dpc-1863	8	47	credited	credit	VERB
dpc-1863	8	48	.	.	PUNCT
dpc-1863	9	1	funding	funding	NOUN
dpc-1863	9	2	:	:	PUNCT
dpc-1863	9	3	none	none	NOUN
dpc-1863	9	4	.	.	PUNCT
dpc-1863	10	1	competing	compete	VERB
dpc-1863	10	2	interests	interest	NOUN
dpc-1863	10	3	:	:	PUNCT
dpc-1863	10	4	pt	pt	PROPN
dpc-1863	10	5	reports	report	NOUN
dpc-1863	10	6	consulting	consult	VERB
dpc-1863	10	7	fees	fee	NOUN
dpc-1863	10	8	from	from	ADP
dpc-1863	10	9	silverchair	silverchair	PROPN
dpc-1863	10	10	,	,	PUNCT
dpc-1863	10	11	honoraria	honorarium	NOUN
dpc-1863	10	12	from	from	ADP
dpc-1863	10	13	fotofinder	fotofinder	NOUN
dpc-1863	10	14	,	,	PUNCT
dpc-1863	10	15	novartis	novartis	NOUN
dpc-1863	10	16	and	and	CCONJ
dpc-1863	10	17	lilly	lilly	PROPN
dpc-1863	10	18	,	,	PUNCT
dpc-1863	10	19	and	and	CCONJ
dpc-1863	10	20	grants	grant	NOUN
dpc-1863	10	21	from	from	ADP
dpc-1863	10	22	metaoptima	metaoptima	PROPN
dpc-1863	10	23	technology	technology	NOUN
dpc-1863	10	24	and	and	CCONJ
dpc-1863	10	25	lilly	lilly	PROPN
dpc-1863	10	26	,	,	PUNCT
dpc-1863	10	27	outside	outside	ADP
dpc-1863	10	28	the	the	DET
dpc-1863	10	29	submitted	submit	VERB
dpc-1863	10	30	work	work	NOUN
dpc-1863	10	31	.	.	PUNCT
dpc-1863	11	1	all	all	DET
dpc-1863	11	2	other	other	ADJ
dpc-1863	11	3	authors	author	NOUN
dpc-1863	11	4	declare	declare	VERB
dpc-1863	11	5	no	no	DET
dpc-1863	11	6	conflict	conflict	NOUN
dpc-1863	11	7	of	of	ADP
dpc-1863	11	8	interest	interest	NOUN
dpc-1863	11	9	.	.	PUNCT
dpc-1863	12	1	authorship	authorship	NOUN
dpc-1863	12	2	:	:	PUNCT
dpc-1863	12	3	all	all	DET
dpc-1863	12	4	authors	author	NOUN
dpc-1863	12	5	have	have	AUX
dpc-1863	12	6	contributed	contribute	VERB
dpc-1863	12	7	significantly	significantly	ADV
dpc-1863	12	8	to	to	ADP
dpc-1863	12	9	this	this	DET
dpc-1863	12	10	publication	publication	NOUN
dpc-1863	12	11	.	.	PUNCT
dpc-1863	13	1	corresponding	correspond	VERB
dpc-1863	13	2	author	author	NOUN
dpc-1863	13	3	:	:	PUNCT
dpc-1863	13	4	philipp	philipp	PROPN
dpc-1863	13	5	tschandl	tschandl	NOUN
dpc-1863	13	6	,	,	PUNCT
dpc-1863	13	7	phd	phd	NOUN
dpc-1863	13	8	,	,	PUNCT
dpc-1863	13	9	department	department	NOUN
dpc-1863	13	10	of	of	ADP
dpc-1863	13	11	dermatology	dermatology	NOUN
dpc-1863	13	12	,	,	PUNCT
dpc-1863	13	13	medical	medical	ADJ
dpc-1863	13	14	university	university	PROPN
dpc-1863	13	15	of	of	ADP
dpc-1863	13	16	vienna	vienna	PROPN
dpc-1863	13	17	,	,	PUNCT
dpc-1863	13	18	währinger	währinger	NOUN
dpc-1863	13	19	gürtel	gürtel	NOUN
dpc-1863	13	20	18	18	NUM
dpc-1863	13	21	-	-	SYM
dpc-1863	13	22	20	20	NUM
dpc-1863	13	23	,	,	PUNCT
dpc-1863	13	24	1090	1090	NUM
dpc-1863	13	25	vienna	vienna	NOUN
dpc-1863	13	26	,	,	PUNCT
dpc-1863	13	27	austria	austria	PROPN
dpc-1863	13	28	,	,	PUNCT
dpc-1863	13	29	e	e	NOUN
dpc-1863	13	30	-	-	NOUN
dpc-1863	13	31	mail	mail	NOUN
dpc-1863	13	32	:	:	PUNCT
dpc-1863	13	33	philipp.tschandl@meduniwien.ac.at	philipp.tschandl@meduniwien.ac.at	NUM
dpc-1863	13	34	introduction	introduction	NOUN
dpc-1863	13	35	:	:	PUNCT
dpc-1863	13	36	classification	classification	NOUN
dpc-1863	13	37	of	of	ADP
dpc-1863	13	38	dermatoscopic	dermatoscopic	ADJ
dpc-1863	13	39	images	image	NOUN
dpc-1863	13	40	via	via	ADP
dpc-1863	13	41	neural	neural	ADJ
dpc-1863	13	42	networks	network	NOUN
dpc-1863	13	43	shows	show	VERB
dpc-1863	13	44	comparable	comparable	ADJ
dpc-1863	13	45	performance	performance	NOUN
dpc-1863	13	46	to	to	ADP
dpc-1863	13	47	clinicians	clinician	NOUN
dpc-1863	13	48	in	in	ADP
dpc-1863	13	49	experimental	experimental	ADJ
dpc-1863	13	50	conditions	condition	NOUN
dpc-1863	13	51	but	but	CCONJ
dpc-1863	13	52	can	can	AUX
dpc-1863	13	53	be	be	AUX
dpc-1863	13	54	affected	affect	VERB
dpc-1863	13	55	by	by	ADP
dpc-1863	13	56	artefacts	artefact	NOUN
dpc-1863	13	57	like	like	ADP
dpc-1863	13	58	skin	skin	NOUN
dpc-1863	13	59	markings	marking	NOUN
dpc-1863	13	60	or	or	CCONJ
dpc-1863	13	61	rulers	ruler	NOUN
dpc-1863	13	62	.	.	PUNCT
dpc-1863	14	1	it	it	PRON
dpc-1863	14	2	is	be	AUX
dpc-1863	14	3	unknown	unknown	ADJ
dpc-1863	14	4	whether	whether	SCONJ
dpc-1863	14	5	specialized	specialized	ADJ
dpc-1863	14	6	neural	neural	ADJ
dpc-1863	14	7	networks	network	NOUN
dpc-1863	14	8	are	be	AUX
dpc-1863	14	9	more	more	ADV
dpc-1863	14	10	robust	robust	ADJ
dpc-1863	14	11	to	to	ADP
dpc-1863	14	12	artefacts	artefact	NOUN
dpc-1863	14	13	.	.	PUNCT
dpc-1863	15	1	objectives	objective	NOUN
dpc-1863	15	2	:	:	PUNCT
dpc-1863	15	3	analyze	analyze	VERB
dpc-1863	15	4	robustness	robustness	NOUN
dpc-1863	15	5	of	of	ADP
dpc-1863	15	6	3	3	NUM
dpc-1863	15	7	neural	neural	ADJ
dpc-1863	15	8	network	network	NOUN
dpc-1863	15	9	architectures	architecture	NOUN
dpc-1863	15	10	,	,	PUNCT
dpc-1863	15	11	namely	namely	ADV
dpc-1863	15	12	resnet-34	resnet-34	PROPN
dpc-1863	15	13	,	,	PUNCT
dpc-1863	15	14	faster	fast	ADJ
dpc-1863	15	15	r	r	NOUN
dpc-1863	15	16	-	-	PUNCT
dpc-1863	15	17	cnn	cnn	PROPN
dpc-1863	15	18	and	and	CCONJ
dpc-1863	15	19	mask	mask	VERB
dpc-1863	15	20	r	r	PROPN
dpc-1863	15	21	-	-	PUNCT
dpc-1863	15	22	cnn	cnn	PROPN
dpc-1863	15	23	.	.	PUNCT
dpc-1863	16	1	methods	method	NOUN
dpc-1863	16	2	:	:	PUNCT
dpc-1863	16	3	we	we	PRON
dpc-1863	16	4	identified	identify	VERB
dpc-1863	16	5	common	common	ADJ
dpc-1863	16	6	artefacts	artefact	NOUN
dpc-1863	16	7	in	in	ADP
dpc-1863	16	8	the	the	DET
dpc-1863	16	9	ham10000	ham10000	NOUN
dpc-1863	16	10	,	,	PUNCT
dpc-1863	16	11	ph2	ph2	NOUN
dpc-1863	16	12	and	and	CCONJ
dpc-1863	16	13	the	the	DET
dpc-1863	16	14	7	7	NUM
dpc-1863	16	15	-	-	PUNCT
dpc-1863	16	16	point	point	NOUN
dpc-1863	16	17	criteria	criterion	NOUN
dpc-1863	16	18	evaluation	evaluation	NOUN
dpc-1863	16	19	datasets	dataset	NOUN
dpc-1863	16	20	,	,	PUNCT
dpc-1863	16	21	and	and	CCONJ
dpc-1863	16	22	established	establish	VERB
dpc-1863	16	23	a	a	DET
dpc-1863	16	24	template	template	NOUN
dpc-1863	16	25	-	-	PUNCT
dpc-1863	16	26	based	base	VERB
dpc-1863	16	27	method	method	NOUN
dpc-1863	16	28	to	to	PART
dpc-1863	16	29	superimpose	superimpose	VERB
dpc-1863	16	30	artefacts	artefact	NOUN
dpc-1863	16	31	on	on	ADP
dpc-1863	16	32	dermatoscopic	dermatoscopic	NOUN
dpc-1863	16	33	images	image	NOUN
dpc-1863	16	34	.	.	PUNCT
dpc-1863	17	1	the	the	DET
dpc-1863	17	2	ham10000	ham10000	ADV
dpc-1863	17	3	-	-	PUNCT
dpc-1863	17	4	dataset	dataset	NOUN
dpc-1863	17	5	with	with	ADP
dpc-1863	17	6	and	and	CCONJ
dpc-1863	17	7	without	without	ADP
dpc-1863	17	8	superimposed	superimposed	ADJ
dpc-1863	17	9	artefacts	artefact	NOUN
dpc-1863	17	10	was	be	AUX
dpc-1863	17	11	used	use	VERB
dpc-1863	17	12	to	to	PART
dpc-1863	17	13	train	train	VERB
dpc-1863	17	14	the	the	DET
dpc-1863	17	15	networks	network	NOUN
dpc-1863	17	16	,	,	PUNCT
dpc-1863	17	17	followed	follow	VERB
dpc-1863	17	18	by	by	ADP
dpc-1863	17	19	analyzing	analyze	VERB
dpc-1863	17	20	their	their	PRON
dpc-1863	17	21	robustness	robustness	NOUN
dpc-1863	17	22	against	against	ADP
dpc-1863	17	23	artefacts	artefact	NOUN
dpc-1863	17	24	in	in	ADP
dpc-1863	17	25	test	test	NOUN
dpc-1863	17	26	images	image	NOUN
dpc-1863	17	27	.	.	PUNCT
dpc-1863	18	1	performance	performance	NOUN
dpc-1863	18	2	was	be	AUX
dpc-1863	18	3	assessed	assess	VERB
dpc-1863	18	4	via	via	ADP
dpc-1863	18	5	area	area	NOUN
dpc-1863	18	6	under	under	ADP
dpc-1863	18	7	the	the	DET
dpc-1863	18	8	precision	precision	NOUN
dpc-1863	18	9	recall	recall	NOUN
dpc-1863	18	10	curve	curve	NOUN
dpc-1863	18	11	and	and	CCONJ
dpc-1863	18	12	classification	classification	NOUN
dpc-1863	18	13	results	result	NOUN
dpc-1863	18	14	.	.	PUNCT
dpc-1863	19	1	results	result	NOUN
dpc-1863	19	2	:	:	PUNCT
dpc-1863	19	3	resnet-34	resnet-34	PROPN
dpc-1863	19	4	and	and	CCONJ
dpc-1863	19	5	faster	fast	ADJ
dpc-1863	19	6	r	r	NOUN
dpc-1863	19	7	-	-	PUNCT
dpc-1863	19	8	cnn	cnn	PROPN
dpc-1863	19	9	models	model	NOUN
dpc-1863	19	10	trained	train	VERB
dpc-1863	19	11	on	on	ADP
dpc-1863	19	12	regular	regular	ADJ
dpc-1863	19	13	images	image	NOUN
dpc-1863	19	14	perform	perform	VERB
dpc-1863	19	15	worse	bad	ADJ
dpc-1863	19	16	than	than	ADP
dpc-1863	19	17	mask	mask	VERB
dpc-1863	19	18	r	r	NOUN
dpc-1863	19	19	-	-	PUNCT
dpc-1863	19	20	cnn	cnn	NOUN
dpc-1863	19	21	on	on	ADP
dpc-1863	19	22	images	image	NOUN
dpc-1863	19	23	with	with	ADP
dpc-1863	19	24	superimposed	superimpose	VERB
dpc-1863	19	25	artefacts	artefact	NOUN
dpc-1863	19	26	.	.	PUNCT
dpc-1863	20	1	artefacts	artefact	NOUN
dpc-1863	20	2	added	add	VERB
dpc-1863	20	3	to	to	ADP
dpc-1863	20	4	all	all	DET
dpc-1863	20	5	tested	test	VERB
dpc-1863	20	6	images	image	NOUN
dpc-1863	20	7	led	lead	VERB
dpc-1863	20	8	to	to	ADP
dpc-1863	20	9	a	a	DET
dpc-1863	20	10	decrease	decrease	NOUN
dpc-1863	20	11	in	in	ADP
dpc-1863	20	12	area	area	NOUN
dpc-1863	20	13	under	under	ADP
dpc-1863	20	14	the	the	DET
dpc-1863	20	15	precision	precision	NOUN
dpc-1863	20	16	-	-	PUNCT
dpc-1863	20	17	recall	recall	NOUN
dpc-1863	20	18	curve	curve	NOUN
dpc-1863	20	19	values	value	NOUN
dpc-1863	20	20	of	of	ADP
dpc-1863	20	21	0.030	0.030	NUM
dpc-1863	20	22	for	for	ADP
dpc-1863	20	23	resnet-34	resnet-34	PROPN
dpc-1863	20	24	and	and	CCONJ
dpc-1863	20	25	0.045	0.045	NUM
dpc-1863	20	26	for	for	ADP
dpc-1863	20	27	faster	fast	ADJ
dpc-1863	20	28	r	r	NOUN
dpc-1863	20	29	-	-	PUNCT
dpc-1863	20	30	cnn	cnn	NOUN
dpc-1863	20	31	in	in	ADP
dpc-1863	20	32	comparison	comparison	NOUN
dpc-1863	20	33	to	to	ADP
dpc-1863	20	34	only	only	ADV
dpc-1863	20	35	0.011	0.011	NUM
dpc-1863	20	36	for	for	ADP
dpc-1863	20	37	mask	mask	NOUN
dpc-1863	20	38	r	r	NOUN
dpc-1863	20	39	-	-	PUNCT
dpc-1863	20	40	cnn	cnn	PROPN
dpc-1863	20	41	.	.	PUNCT
dpc-1863	21	1	however	however	ADV
dpc-1863	21	2	,	,	PUNCT
dpc-1863	21	3	changes	change	NOUN
dpc-1863	21	4	in	in	ADP
dpc-1863	21	5	model	model	NOUN
dpc-1863	21	6	performance	performance	NOUN
dpc-1863	21	7	only	only	ADV
dpc-1863	21	8	became	become	VERB
dpc-1863	21	9	significant	significant	ADJ
dpc-1863	21	10	with	with	ADP
dpc-1863	21	11	40	40	NUM
dpc-1863	21	12	%	%	NOUN
dpc-1863	21	13	or	or	CCONJ
dpc-1863	21	14	more	more	ADJ
dpc-1863	21	15	of	of	ADP
dpc-1863	21	16	the	the	DET
dpc-1863	21	17	images	image	NOUN
dpc-1863	21	18	having	have	VERB
dpc-1863	21	19	superimposed	superimpose	VERB
dpc-1863	21	20	artefacts	artefact	NOUN
dpc-1863	21	21	.	.	PUNCT
dpc-1863	22	1	a	a	DET
dpc-1863	22	2	loss	loss	NOUN
dpc-1863	22	3	in	in	ADP
dpc-1863	22	4	performance	performance	NOUN
dpc-1863	22	5	occurred	occur	VERB
dpc-1863	22	6	when	when	SCONJ
dpc-1863	22	7	the	the	DET
dpc-1863	22	8	training	training	NOUN
dpc-1863	22	9	was	be	AUX
dpc-1863	22	10	biased	bias	VERB
dpc-1863	22	11	by	by	ADP
dpc-1863	22	12	selectively	selectively	ADV
dpc-1863	22	13	superimposing	superimpose	VERB
dpc-1863	22	14	artefacts	artefact	NOUN
dpc-1863	22	15	on	on	ADP
dpc-1863	22	16	images	image	NOUN
dpc-1863	22	17	belonging	belong	VERB
dpc-1863	22	18	to	to	ADP
dpc-1863	22	19	a	a	DET
dpc-1863	22	20	certain	certain	ADJ
dpc-1863	22	21	class	class	NOUN
dpc-1863	22	22	.	.	PUNCT
dpc-1863	23	1	conclusions	conclusion	NOUN
dpc-1863	23	2	:	:	PUNCT
dpc-1863	23	3	as	as	SCONJ
dpc-1863	23	4	mask	mask	NOUN
dpc-1863	23	5	r	r	NOUN
dpc-1863	23	6	-	-	PUNCT
dpc-1863	23	7	cnn	cnn	PROPN
dpc-1863	23	8	showed	show	VERB
dpc-1863	23	9	the	the	DET
dpc-1863	23	10	least	least	ADJ
dpc-1863	23	11	decrease	decrease	NOUN
dpc-1863	23	12	in	in	ADP
dpc-1863	23	13	performance	performance	NOUN
dpc-1863	23	14	when	when	SCONJ
dpc-1863	23	15	confronted	confront	VERB
dpc-1863	23	16	with	with	ADP
dpc-1863	23	17	artefacts	artefact	NOUN
dpc-1863	23	18	,	,	PUNCT
dpc-1863	23	19	instance	instance	NOUN
dpc-1863	23	20	segmentation	segmentation	NOUN
dpc-1863	23	21	architectures	architecture	NOUN
dpc-1863	23	22	may	may	AUX
dpc-1863	23	23	be	be	AUX
dpc-1863	23	24	helpful	helpful	ADJ
dpc-1863	23	25	to	to	PART
dpc-1863	23	26	counter	counter	VERB
dpc-1863	23	27	the	the	DET
dpc-1863	23	28	effects	effect	NOUN
dpc-1863	23	29	of	of	ADP
dpc-1863	23	30	artefacts	artefact	NOUN
dpc-1863	23	31	,	,	PUNCT
dpc-1863	23	32	warranting	warrant	VERB
dpc-1863	23	33	further	further	ADJ
dpc-1863	23	34	research	research	NOUN
dpc-1863	23	35	on	on	ADP
dpc-1863	23	36	related	related	ADJ
dpc-1863	23	37	architectures	architecture	NOUN
dpc-1863	23	38	.	.	PUNCT
dpc-1863	24	1	our	our	PRON
dpc-1863	24	2	artefact	artefact	ADJ
dpc-1863	24	3	insertion	insertion	NOUN
dpc-1863	24	4	mechanism	mechanism	NOUN
dpc-1863	24	5	could	could	AUX
dpc-1863	24	6	be	be	AUX
dpc-1863	24	7	useful	useful	ADJ
dpc-1863	24	8	for	for	ADP
dpc-1863	24	9	future	future	ADJ
dpc-1863	24	10	research	research	NOUN
dpc-1863	24	11	.	.	PUNCT
dpc-1863	25	1	abstract	abstract	ADJ
dpc-1863	25	2	2	2	NUM
dpc-1863	25	3	original	original	ADJ
dpc-1863	25	4	article	article	NOUN
dpc-1863	25	5	|	|	ADV
dpc-1863	25	6	dermatol	dermatol	ADJ
dpc-1863	25	7	pract	pract	NOUN
dpc-1863	25	8	concept	concept	NOUN
dpc-1863	25	9	.	.	PUNCT
dpc-1863	26	1	2022;12(3):e2022126	2022;12(3):e2022126	NUM
dpc-1863	26	2	introduction	introduction	NOUN
dpc-1863	26	3	epidemiological	epidemiological	ADJ
dpc-1863	26	4	studies	study	NOUN
dpc-1863	26	5	show	show	VERB
dpc-1863	26	6	an	an	DET
dpc-1863	26	7	increasing	increase	VERB
dpc-1863	26	8	trend	trend	NOUN
dpc-1863	26	9	in	in	ADP
dpc-1863	26	10	the	the	DET
dpc-1863	26	11	incidence	incidence	NOUN
dpc-1863	26	12	rates	rate	NOUN
dpc-1863	26	13	of	of	ADP
dpc-1863	26	14	melanoma	melanoma	NOUN
dpc-1863	26	15	and	and	CCONJ
dpc-1863	26	16	non	non	ADJ
dpc-1863	26	17	-	-	ADJ
dpc-1863	26	18	melanoma	melanoma	ADJ
dpc-1863	26	19	skin	skin	NOUN
dpc-1863	26	20	cancer	cancer	NOUN
dpc-1863	26	21	worldwide	worldwide	ADV
dpc-1863	26	22	over	over	ADP
dpc-1863	26	23	the	the	DET
dpc-1863	26	24	last	last	ADJ
dpc-1863	26	25	30	30	NUM
dpc-1863	26	26	years	year	NOUN
dpc-1863	26	27	[	[	X
dpc-1863	26	28	1	1	NUM
dpc-1863	26	29	]	]	PUNCT
dpc-1863	26	30	.	.	PUNCT
dpc-1863	27	1	according	accord	VERB
dpc-1863	27	2	to	to	ADP
dpc-1863	27	3	the	the	DET
dpc-1863	27	4	american	american	PROPN
dpc-1863	27	5	joint	joint	PROPN
dpc-1863	27	6	committee	committee	PROPN
dpc-1863	27	7	on	on	ADP
dpc-1863	27	8	cancer	cancer	NOUN
dpc-1863	27	9	melanoma	melanoma	NOUN
dpc-1863	27	10	staging	staging	NOUN
dpc-1863	27	11	system	system	NOUN
dpc-1863	27	12	,	,	PUNCT
dpc-1863	27	13	stage	stage	NOUN
dpc-1863	27	14	i	i	PRON
dpc-1863	27	15	malignant	malignant	VERB
dpc-1863	27	16	skin	skin	NOUN
dpc-1863	27	17	alterations	alteration	NOUN
dpc-1863	27	18	with	with	ADP
dpc-1863	27	19	a	a	DET
dpc-1863	27	20	five	five	NUM
dpc-1863	27	21	-	-	PUNCT
dpc-1863	27	22	year	year	NOUN
dpc-1863	27	23	survival	survival	NOUN
dpc-1863	27	24	rate	rate	NOUN
dpc-1863	27	25	of	of	ADP
dpc-1863	27	26	more	more	ADJ
dpc-1863	27	27	than	than	ADP
dpc-1863	27	28	90	90	NUM
dpc-1863	27	29	%	%	NOUN
dpc-1863	27	30	contrasts	contrast	VERB
dpc-1863	27	31	with	with	ADP
dpc-1863	27	32	a	a	DET
dpc-1863	27	33	survival	survival	NOUN
dpc-1863	27	34	rate	rate	NOUN
dpc-1863	27	35	of	of	ADP
dpc-1863	27	36	less	less	ADJ
dpc-1863	27	37	than	than	ADP
dpc-1863	27	38	15	15	NUM
dpc-1863	27	39	%	%	NOUN
dpc-1863	27	40	for	for	ADP
dpc-1863	27	41	stage	stage	NOUN
dpc-1863	27	42	iv	iv	NUM
dpc-1863	27	43	patients	patient	NOUN
dpc-1863	27	44	.	.	PUNCT
dpc-1863	28	1	this	this	PRON
dpc-1863	28	2	indicates	indicate	VERB
dpc-1863	28	3	a	a	DET
dpc-1863	28	4	clear	clear	ADJ
dpc-1863	28	5	need	need	NOUN
dpc-1863	28	6	for	for	ADP
dpc-1863	28	7	early	early	ADJ
dpc-1863	28	8	,	,	PUNCT
dpc-1863	28	9	reliable	reliable	ADJ
dpc-1863	28	10	and	and	CCONJ
dpc-1863	28	11	consistent	consistent	ADJ
dpc-1863	28	12	diagnosis	diagnosis	NOUN
dpc-1863	28	13	and	and	CCONJ
dpc-1863	28	14	treatment	treatment	NOUN
dpc-1863	28	15	[	[	X
dpc-1863	28	16	2	2	NUM
dpc-1863	28	17	]	]	PUNCT
dpc-1863	28	18	.	.	PUNCT
dpc-1863	29	1	the	the	DET
dpc-1863	29	2	desire	desire	NOUN
dpc-1863	29	3	for	for	ADP
dpc-1863	29	4	automatic	automatic	ADJ
dpc-1863	29	5	lesion	lesion	NOUN
dpc-1863	29	6	analysis	analysis	NOUN
dpc-1863	29	7	is	be	AUX
dpc-1863	29	8	further	far	ADV
dpc-1863	29	9	intensified	intensify	VERB
dpc-1863	29	10	by	by	ADP
dpc-1863	29	11	a	a	DET
dpc-1863	29	12	high	high	ADJ
dpc-1863	29	13	dependency	dependency	NOUN
dpc-1863	29	14	between	between	ADP
dpc-1863	29	15	the	the	DET
dpc-1863	29	16	diagnostic	diagnostic	ADJ
dpc-1863	29	17	quality	quality	NOUN
dpc-1863	29	18	and	and	CCONJ
dpc-1863	29	19	the	the	DET
dpc-1863	29	20	examiners	examiner	NOUN
dpc-1863	29	21	experience	experience	VERB
dpc-1863	29	22	in	in	ADP
dpc-1863	29	23	dermoscopy	dermoscopy	NOUN
dpc-1863	29	24	,	,	PUNCT
dpc-1863	29	25	as	as	ADV
dpc-1863	29	26	well	well	ADV
dpc-1863	29	27	as	as	ADP
dpc-1863	29	28	a	a	DET
dpc-1863	29	29	high	high	ADJ
dpc-1863	29	30	degree	degree	NOUN
dpc-1863	29	31	of	of	ADP
dpc-1863	29	32	interand	interand	NOUN
dpc-1863	29	33	intra	intra	ADJ
dpc-1863	29	34	-	-	NOUN
dpc-1863	29	35	variability	variability	NOUN
dpc-1863	29	36	of	of	ADP
dpc-1863	29	37	diagnoses	diagnosis	NOUN
dpc-1863	29	38	[	[	X
dpc-1863	29	39	3,4	3,4	NUM
dpc-1863	29	40	]	]	PUNCT
dpc-1863	29	41	.	.	PUNCT
dpc-1863	30	1	methods	method	NOUN
dpc-1863	30	2	of	of	ADP
dpc-1863	30	3	automatic	automatic	ADJ
dpc-1863	30	4	skin	skin	NOUN
dpc-1863	30	5	lesion	lesion	NOUN
dpc-1863	30	6	analysis	analysis	NOUN
dpc-1863	30	7	have	have	AUX
dpc-1863	30	8	been	be	AUX
dpc-1863	30	9	the	the	DET
dpc-1863	30	10	focus	focus	NOUN
dpc-1863	30	11	of	of	ADP
dpc-1863	30	12	research	research	NOUN
dpc-1863	30	13	for	for	ADP
dpc-1863	30	14	decades	decade	NOUN
dpc-1863	30	15	,	,	PUNCT
dpc-1863	30	16	and	and	CCONJ
dpc-1863	30	17	have	have	AUX
dpc-1863	30	18	gained	gain	VERB
dpc-1863	30	19	interest	interest	NOUN
dpc-1863	30	20	in	in	ADP
dpc-1863	30	21	recent	recent	ADJ
dpc-1863	30	22	years	year	NOUN
dpc-1863	31	1	[	[	X
dpc-1863	31	2	5,6	5,6	NUM
dpc-1863	31	3	]	]	PUNCT
dpc-1863	31	4	.	.	PUNCT
dpc-1863	32	1	these	these	DET
dpc-1863	32	2	methods	method	NOUN
dpc-1863	32	3	are	be	AUX
dpc-1863	32	4	intended	intend	VERB
dpc-1863	32	5	to	to	PART
dpc-1863	32	6	support	support	VERB
dpc-1863	32	7	tele	tele	NOUN
dpc-1863	32	8	-	-	PUNCT
dpc-1863	32	9	dermatologic	dermatologic	ADJ
dpc-1863	32	10	settings	setting	NOUN
dpc-1863	32	11	,	,	PUNCT
dpc-1863	32	12	improve	improve	VERB
dpc-1863	32	13	management	management	NOUN
dpc-1863	32	14	decisions	decision	NOUN
dpc-1863	32	15	or	or	CCONJ
dpc-1863	32	16	aid	aid	NOUN
dpc-1863	32	17	in	in	ADP
dpc-1863	32	18	difficult	difficult	ADJ
dpc-1863	32	19	clinical	clinical	ADJ
dpc-1863	32	20	scenarios	scenario	NOUN
dpc-1863	32	21	,	,	PUNCT
dpc-1863	32	22	but	but	CCONJ
dpc-1863	32	23	often	often	ADV
dpc-1863	32	24	suffer	suffer	VERB
dpc-1863	32	25	,	,	PUNCT
dpc-1863	32	26	among	among	ADP
dpc-1863	32	27	other	other	ADJ
dpc-1863	32	28	things	thing	NOUN
dpc-1863	32	29	,	,	PUNCT
dpc-1863	32	30	from	from	ADP
dpc-1863	32	31	the	the	DET
dpc-1863	32	32	presence	presence	NOUN
dpc-1863	32	33	of	of	ADP
dpc-1863	32	34	artefacts	artefact	NOUN
dpc-1863	32	35	in	in	ADP
dpc-1863	32	36	dermatoscopic	dermatoscopic	NOUN
dpc-1863	32	37	images	image	NOUN
dpc-1863	32	38	[	[	X
dpc-1863	32	39	7	7	NUM
dpc-1863	32	40	-	-	SYM
dpc-1863	32	41	12	12	NUM
dpc-1863	32	42	]	]	PUNCT
dpc-1863	32	43	.	.	PUNCT
dpc-1863	33	1	a	a	DET
dpc-1863	33	2	common	common	ADJ
dpc-1863	33	3	neural	neural	ADJ
dpc-1863	33	4	network	network	NOUN
dpc-1863	33	5	used	use	VERB
dpc-1863	33	6	for	for	ADP
dpc-1863	33	7	classification	classification	NOUN
dpc-1863	33	8	is	be	AUX
dpc-1863	33	9	resnet	resnet	ADJ
dpc-1863	33	10	,	,	PUNCT
dpc-1863	33	11	two	two	NUM
dpc-1863	33	12	well	well	ADV
dpc-1863	33	13	-	-	PUNCT
dpc-1863	33	14	known	know	VERB
dpc-1863	33	15	neural	neural	ADJ
dpc-1863	33	16	network	network	NOUN
dpc-1863	33	17	architectures	architecture	NOUN
dpc-1863	33	18	in	in	ADP
dpc-1863	33	19	computer	computer	NOUN
dpc-1863	33	20	vision	vision	NOUN
dpc-1863	33	21	are	be	AUX
dpc-1863	33	22	faster	fast	ADJ
dpc-1863	33	23	r	r	X
dpc-1863	33	24	-	-	PUNCT
dpc-1863	33	25	cnn	cnn	PROPN
dpc-1863	33	26	and	and	CCONJ
dpc-1863	33	27	mask	mask	VERB
dpc-1863	33	28	r	r	PROPN
dpc-1863	33	29	-	-	PUNCT
dpc-1863	33	30	cnn	cnn	PROPN
dpc-1863	33	31	(	(	PUNCT
dpc-1863	33	32	figure	figure	NOUN
dpc-1863	33	33	1	1	NUM
dpc-1863	33	34	)	)	PUNCT
dpc-1863	34	1	[	[	X
dpc-1863	34	2	13,14	13,14	NUM
dpc-1863	34	3	]	]	PUNCT
dpc-1863	34	4	.	.	PUNCT
dpc-1863	35	1	the	the	DET
dpc-1863	35	2	first	first	ADJ
dpc-1863	35	3	is	be	AUX
dpc-1863	35	4	performing	perform	VERB
dpc-1863	35	5	“	"	PUNCT
dpc-1863	35	6	object	object	NOUN
dpc-1863	35	7	detection	detection	NOUN
dpc-1863	35	8	”	"	PUNCT
dpc-1863	35	9	,	,	PUNCT
dpc-1863	35	10	a	a	DET
dpc-1863	35	11	process	process	NOUN
dpc-1863	35	12	where	where	SCONJ
dpc-1863	35	13	one	one	NUM
dpc-1863	35	14	or	or	CCONJ
dpc-1863	35	15	multiple	multiple	ADJ
dpc-1863	35	16	objects	object	NOUN
dpc-1863	35	17	in	in	ADP
dpc-1863	35	18	an	an	DET
dpc-1863	35	19	image	image	NOUN
dpc-1863	35	20	can	can	AUX
dpc-1863	35	21	be	be	AUX
dpc-1863	35	22	detected	detect	VERB
dpc-1863	35	23	and	and	CCONJ
dpc-1863	35	24	located	locate	VERB
dpc-1863	35	25	with	with	ADP
dpc-1863	35	26	a	a	DET
dpc-1863	35	27	rectangular	rectangular	ADJ
dpc-1863	35	28	“	"	PUNCT
dpc-1863	35	29	bounding	bounding	NOUN
dpc-1863	35	30	box	box	NOUN
dpc-1863	35	31	”	"	PUNCT
dpc-1863	35	32	.	.	PUNCT
dpc-1863	36	1	the	the	DET
dpc-1863	36	2	latter	latter	ADJ
dpc-1863	36	3	is	be	AUX
dpc-1863	36	4	performing	perform	VERB
dpc-1863	36	5	“	"	PUNCT
dpc-1863	36	6	instance	instance	NOUN
dpc-1863	36	7	segmentation	segmentation	NOUN
dpc-1863	36	8	”	"	PUNCT
dpc-1863	36	9	where	where	SCONJ
dpc-1863	36	10	one	one	NUM
dpc-1863	36	11	or	or	CCONJ
dpc-1863	36	12	more	more	ADJ
dpc-1863	36	13	objects	object	NOUN
dpc-1863	36	14	in	in	ADP
dpc-1863	36	15	an	an	DET
dpc-1863	36	16	image	image	NOUN
dpc-1863	36	17	can	can	AUX
dpc-1863	36	18	be	be	AUX
dpc-1863	36	19	found	find	VERB
dpc-1863	36	20	and	and	CCONJ
dpc-1863	36	21	their	their	PRON
dpc-1863	36	22	respective	respective	ADJ
dpc-1863	36	23	area	area	NOUN
dpc-1863	36	24	(	(	PUNCT
dpc-1863	36	25	i.e.	i.e.	X
dpc-1863	36	26	pixels	pixel	NOUN
dpc-1863	36	27	)	)	PUNCT
dpc-1863	36	28	in	in	ADP
dpc-1863	36	29	the	the	DET
dpc-1863	36	30	image	image	NOUN
dpc-1863	36	31	outlined	outline	VERB
dpc-1863	36	32	(	(	PUNCT
dpc-1863	36	33	“	"	PUNCT
dpc-1863	36	34	segmented	segment	VERB
dpc-1863	36	35	”	"	PUNCT
dpc-1863	36	36	)	)	PUNCT
dpc-1863	36	37	,	,	PUNCT
dpc-1863	36	38	and	and	CCONJ
dpc-1863	36	39	can	can	AUX
dpc-1863	36	40	be	be	AUX
dpc-1863	36	41	regarded	regard	VERB
dpc-1863	36	42	as	as	ADP
dpc-1863	36	43	a	a	DET
dpc-1863	36	44	cnn	cnn	PROPN
dpc-1863	36	45	-	-	PUNCT
dpc-1863	36	46	based	base	VERB
dpc-1863	36	47	multi	multi	ADJ
dpc-1863	36	48	-	-	ADJ
dpc-1863	36	49	instance	instance	NOUN
dpc-1863	36	50	generalisation	generalisation	NOUN
dpc-1863	36	51	of	of	ADP
dpc-1863	36	52	computer	computer	NOUN
dpc-1863	36	53	-	-	PUNCT
dpc-1863	36	54	vision	vision	NOUN
dpc-1863	36	55	based	base	VERB
dpc-1863	36	56	techniques	technique	NOUN
dpc-1863	36	57	of	of	ADP
dpc-1863	36	58	lesion	lesion	NOUN
dpc-1863	36	59	segmentation	segmentation	NOUN
dpc-1863	36	60	[	[	X
dpc-1863	36	61	15,16	15,16	NUM
dpc-1863	36	62	]	]	PUNCT
dpc-1863	36	63	.	.	PUNCT
dpc-1863	37	1	object	object	NOUN
dpc-1863	37	2	detection	detection	NOUN
dpc-1863	37	3	has	have	AUX
dpc-1863	37	4	been	be	AUX
dpc-1863	37	5	used	use	VERB
dpc-1863	37	6	in	in	ADP
dpc-1863	37	7	the	the	DET
dpc-1863	37	8	field	field	NOUN
dpc-1863	37	9	of	of	ADP
dpc-1863	37	10	automated	automate	VERB
dpc-1863	37	11	skin	skin	NOUN
dpc-1863	37	12	cancer	cancer	NOUN
dpc-1863	37	13	detection	detection	NOUN
dpc-1863	37	14	on	on	ADP
dpc-1863	37	15	clinical	clinical	ADJ
dpc-1863	37	16	images	image	NOUN
dpc-1863	37	17	[	[	X
dpc-1863	37	18	17	17	NUM
dpc-1863	37	19	]	]	PUNCT
dpc-1863	37	20	,	,	PUNCT
dpc-1863	37	21	but	but	CCONJ
dpc-1863	37	22	training	training	NOUN
dpc-1863	37	23	of	of	ADP
dpc-1863	37	24	instance	instance	NOUN
dpc-1863	37	25	segmentation	segmentation	NOUN
dpc-1863	37	26	neural	neural	ADJ
dpc-1863	37	27	networks	network	NOUN
dpc-1863	37	28	in	in	ADP
dpc-1863	37	29	dermatoscopy	dermatoscopy	NOUN
dpc-1863	37	30	has	have	AUX
dpc-1863	37	31	not	not	PART
dpc-1863	37	32	yet	yet	ADV
dpc-1863	37	33	been	be	AUX
dpc-1863	37	34	reported	report	VERB
dpc-1863	37	35	on	on	ADP
dpc-1863	37	36	successfully	successfully	ADV
dpc-1863	37	37	,	,	PUNCT
dpc-1863	37	38	most	most	ADV
dpc-1863	37	39	probably	probably	ADV
dpc-1863	37	40	because	because	SCONJ
dpc-1863	37	41	of	of	ADP
dpc-1863	37	42	missing	miss	VERB
dpc-1863	37	43	ground	ground	NOUN
dpc-1863	37	44	-	-	PUNCT
dpc-1863	37	45	truth	truth	NOUN
dpc-1863	37	46	data	datum	NOUN
dpc-1863	37	47	.	.	PUNCT
dpc-1863	38	1	objectives	objective	VERB
dpc-1863	38	2	our	our	PRON
dpc-1863	38	3	hypothesis	hypothesis	NOUN
dpc-1863	38	4	is	be	AUX
dpc-1863	38	5	that	that	SCONJ
dpc-1863	38	6	in	in	ADP
dpc-1863	38	7	contrast	contrast	NOUN
dpc-1863	38	8	to	to	ADP
dpc-1863	38	9	resnet	resnet	NOUN
dpc-1863	38	10	,	,	PUNCT
dpc-1863	38	11	the	the	DET
dpc-1863	38	12	other	other	ADJ
dpc-1863	38	13	network	network	NOUN
dpc-1863	38	14	architectures	architecture	NOUN
dpc-1863	38	15	intrinsically	intrinsically	ADV
dpc-1863	38	16	have	have	VERB
dpc-1863	38	17	to	to	PART
dpc-1863	38	18	“	"	PUNCT
dpc-1863	38	19	concentrate	concentrate	VERB
dpc-1863	38	20	”	"	PUNCT
dpc-1863	38	21	on	on	ADP
dpc-1863	38	22	regions	region	NOUN
dpc-1863	38	23	of	of	ADP
dpc-1863	38	24	the	the	DET
dpc-1863	38	25	classified	classified	ADJ
dpc-1863	38	26	object	object	NOUN
dpc-1863	38	27	in	in	ADP
dpc-1863	38	28	an	an	DET
dpc-1863	38	29	image	image	NOUN
dpc-1863	38	30	and	and	CCONJ
dpc-1863	38	31	hence	hence	ADV
dpc-1863	38	32	may	may	AUX
dpc-1863	38	33	offer	offer	VERB
dpc-1863	38	34	robustness	robustness	NOUN
dpc-1863	38	35	against	against	ADP
dpc-1863	38	36	artefacts	artefact	NOUN
dpc-1863	38	37	surrounding	surround	VERB
dpc-1863	38	38	the	the	DET
dpc-1863	38	39	lesions	lesion	NOUN
dpc-1863	38	40	.	.	PUNCT
dpc-1863	39	1	robustness	robustness	NOUN
dpc-1863	39	2	in	in	ADP
dpc-1863	39	3	this	this	DET
dpc-1863	39	4	case	case	NOUN
dpc-1863	39	5	describes	describe	VERB
dpc-1863	39	6	the	the	DET
dpc-1863	39	7	consistency	consistency	NOUN
dpc-1863	39	8	of	of	ADP
dpc-1863	39	9	the	the	DET
dpc-1863	39	10	obtained	obtain	VERB
dpc-1863	39	11	diagnoses	diagnosis	NOUN
dpc-1863	39	12	under	under	ADP
dpc-1863	39	13	the	the	DET
dpc-1863	39	14	influence	influence	NOUN
dpc-1863	39	15	of	of	ADP
dpc-1863	39	16	artefacts	artefact	NOUN
dpc-1863	39	17	in	in	ADP
dpc-1863	39	18	the	the	DET
dpc-1863	39	19	input	input	NOUN
dpc-1863	39	20	image	image	NOUN
dpc-1863	39	21	data	datum	NOUN
dpc-1863	39	22	.	.	PUNCT
dpc-1863	40	1	these	these	DET
dpc-1863	40	2	networks	network	NOUN
dpc-1863	40	3	could	could	AUX
dpc-1863	40	4	potentially	potentially	ADV
dpc-1863	40	5	be	be	AUX
dpc-1863	40	6	used	use	VERB
dpc-1863	40	7	as	as	ADP
dpc-1863	40	8	off	off	ADP
dpc-1863	40	9	-	-	PUNCT
dpc-1863	40	10	theshelf	theshelf	NOUN
dpc-1863	40	11	methods	method	NOUN
dpc-1863	40	12	with	with	ADP
dpc-1863	40	13	little	little	ADJ
dpc-1863	40	14	customization	customization	NOUN
dpc-1863	40	15	-	-	PUNCT
dpc-1863	40	16	effort	effort	NOUN
dpc-1863	40	17	needed	need	VERB
dpc-1863	40	18	and	and	CCONJ
dpc-1863	40	19	could	could	AUX
dpc-1863	40	20	enable	enable	VERB
dpc-1863	40	21	us	we	PRON
dpc-1863	40	22	to	to	PART
dpc-1863	40	23	focus	focus	VERB
dpc-1863	40	24	less	less	ADV
dpc-1863	40	25	on	on	ADP
dpc-1863	40	26	tedious	tedious	ADJ
dpc-1863	40	27	image	image	NOUN
dpc-1863	40	28	pre	pre	ADJ
dpc-1863	40	29	-	-	ADJ
dpc-1863	40	30	processing	process	VERB
dpc-1863	40	31	such	such	ADJ
dpc-1863	40	32	as	as	ADP
dpc-1863	40	33	removal	removal	NOUN
dpc-1863	40	34	of	of	ADP
dpc-1863	40	35	bubbles	bubble	NOUN
dpc-1863	40	36	or	or	CCONJ
dpc-1863	40	37	hairs	hair	NOUN
dpc-1863	40	38	[	[	X
dpc-1863	40	39	18	18	NUM
dpc-1863	40	40	]	]	PUNCT
dpc-1863	40	41	.	.	PUNCT
dpc-1863	41	1	methods	method	NOUN
dpc-1863	41	2	image	image	NOUN
dpc-1863	41	3	datasets	dataset	VERB
dpc-1863	41	4	the	the	DET
dpc-1863	41	5	primary	primary	ADJ
dpc-1863	41	6	source	source	NOUN
dpc-1863	41	7	of	of	ADP
dpc-1863	41	8	dermatoscopic	dermatoscopic	ADJ
dpc-1863	41	9	images	image	NOUN
dpc-1863	41	10	was	be	AUX
dpc-1863	41	11	the	the	DET
dpc-1863	41	12	ham10000	ham10000	PROPN
dpc-1863	41	13	dataset	dataset	NOUN
dpc-1863	42	1	[	[	X
dpc-1863	42	2	19	19	NUM
dpc-1863	42	3	]	]	PUNCT
dpc-1863	42	4	.	.	PUNCT
dpc-1863	43	1	this	this	DET
dpc-1863	43	2	dataset	dataset	NOUN
dpc-1863	43	3	also	also	ADV
dpc-1863	43	4	includes	include	VERB
dpc-1863	43	5	publicly	publicly	ADV
dpc-1863	43	6	available	available	ADJ
dpc-1863	43	7	lesion	lesion	NOUN
dpc-1863	43	8	segmentation	segmentation	NOUN
dpc-1863	43	9	masks	mask	NOUN
dpc-1863	43	10	for	for	ADP
dpc-1863	43	11	every	every	DET
dpc-1863	43	12	image	image	NOUN
dpc-1863	43	13	,	,	PUNCT
dpc-1863	43	14	as	as	SCONJ
dpc-1863	43	15	described	describe	VERB
dpc-1863	43	16	previously	previously	ADV
dpc-1863	43	17	,	,	PUNCT
dpc-1863	43	18	which	which	PRON
dpc-1863	43	19	are	be	AUX
dpc-1863	43	20	necessary	necessary	ADJ
dpc-1863	43	21	for	for	ADP
dpc-1863	43	22	training	train	VERB
dpc-1863	43	23	the	the	DET
dpc-1863	43	24	faster	fast	ADJ
dpc-1863	43	25	r	r	NOUN
dpc-1863	43	26	-	-	PUNCT
dpc-1863	43	27	cnn	cnn	PROPN
dpc-1863	43	28	and	and	CCONJ
dpc-1863	43	29	mask	mask	VERB
dpc-1863	43	30	r	r	PROPN
dpc-1863	43	31	-	-	PUNCT
dpc-1863	43	32	cnn	cnn	PROPN
dpc-1863	43	33	architectures	architecture	NOUN
dpc-1863	43	34	[	[	X
dpc-1863	43	35	8	8	NUM
dpc-1863	43	36	]	]	PUNCT
dpc-1863	43	37	.	.	PUNCT
dpc-1863	44	1	it	it	PRON
dpc-1863	44	2	contains	contain	VERB
dpc-1863	44	3	10,015	10,015	NUM
dpc-1863	44	4	images	image	NOUN
dpc-1863	44	5	,	,	PUNCT
dpc-1863	44	6	each	each	PRON
dpc-1863	44	7	with	with	ADP
dpc-1863	44	8	600x450	600x450	NOUN
dpc-1863	44	9	pixels	pixel	NOUN
dpc-1863	44	10	and	and	CCONJ
dpc-1863	44	11	3	3	NUM
dpc-1863	44	12	-	-	SYM
dpc-1863	44	13	8	8	NUM
dpc-1863	44	14	bit	bit	NOUN
dpc-1863	44	15	color	color	NOUN
dpc-1863	44	16	channels	channel	NOUN
dpc-1863	44	17	.	.	PUNCT
dpc-1863	45	1	each	each	DET
dpc-1863	45	2	image	image	NOUN
dpc-1863	45	3	is	be	AUX
dpc-1863	45	4	assigned	assign	VERB
dpc-1863	45	5	one	one	NUM
dpc-1863	45	6	of	of	ADP
dpc-1863	45	7	seven	seven	NUM
dpc-1863	45	8	diagnostic	diagnostic	ADJ
dpc-1863	45	9	classes	class	NOUN
dpc-1863	45	10	:	:	PUNCT
dpc-1863	45	11	actinic	actinic	ADJ
dpc-1863	45	12	keratosis	keratosis	NOUN
dpc-1863	45	13	/	/	SYM
dpc-1863	45	14	intraepithelial	intraepithelial	ADJ
dpc-1863	45	15	carcinoma	carcinoma	NOUN
dpc-1863	45	16	classi	classi	PROPN
dpc-1863	45	17	�	�	PROPN
dpc-1863	45	18	cation	cation	NOUN
dpc-1863	45	19	object	object	NOUN
dpc-1863	45	20	detection	detection	NOUN
dpc-1863	45	21	instance	instance	NOUN
dpc-1863	45	22	segmentation	segmentation	NOUN
dpc-1863	45	23	mask	mask	NOUN
dpc-1863	45	24	r	r	NOUN
dpc-1863	45	25	-	-	PUNCT
dpc-1863	45	26	cnnfaster	cnnfaster	NOUN
dpc-1863	45	27	r	r	NOUN
dpc-1863	45	28	-	-	PUNCT
dpc-1863	45	29	cnnresnet-34	cnnresnet-34	ADJ
dpc-1863	45	30	nevus	nevus	ADJ
dpc-1863	45	31	nevus	nevus	ADJ
dpc-1863	45	32	nevus	nevus	NOUN
dpc-1863	45	33	o	o	PROPN
dpc-1863	45	34	u	u	NOUN
dpc-1863	45	35	tp	tp	ADP
dpc-1863	45	36	u	u	PROPN
dpc-1863	45	37	t	t	PROPN
dpc-1863	46	1	a	a	DET
dpc-1863	46	2	rc	rc	PROPN
dpc-1863	46	3	h	h	NOUN
dpc-1863	47	1	it	it	PRON
dpc-1863	47	2	ec	ec	PROPN
dpc-1863	47	3	tu	tu	INTJ
dpc-1863	47	4	re	re	PROPN
dpc-1863	47	5	in	in	ADP
dpc-1863	47	6	pu	pu	PROPN
dpc-1863	47	7	t	t	PROPN
dpc-1863	47	8	a	a	DET
dpc-1863	47	9	pp	pp	INTJ
dpc-1863	47	10	ro	ro	INTJ
dpc-1863	47	11	a	a	DET
dpc-1863	47	12	ch	ch	NOUN
dpc-1863	47	13	figure	figure	NOUN
dpc-1863	47	14	1	1	NUM
dpc-1863	47	15	.	.	PUNCT
dpc-1863	48	1	a	a	DET
dpc-1863	48	2	visual	visual	ADJ
dpc-1863	48	3	representation	representation	NOUN
dpc-1863	48	4	of	of	ADP
dpc-1863	48	5	the	the	DET
dpc-1863	48	6	outputs	output	NOUN
dpc-1863	48	7	of	of	ADP
dpc-1863	48	8	the	the	DET
dpc-1863	48	9	three	three	NUM
dpc-1863	48	10	approaches	approach	NOUN
dpc-1863	48	11	.	.	PUNCT
dpc-1863	49	1	image	image	NOUN
dpc-1863	49	2	classification	classification	NOUN
dpc-1863	49	3	(	(	PUNCT
dpc-1863	49	4	ie	ie	X
dpc-1863	49	5	resnet-34	resnet-34	PROPN
dpc-1863	49	6	)	)	PUNCT
dpc-1863	49	7	classifies	classify	VERB
dpc-1863	49	8	the	the	DET
dpc-1863	49	9	image	image	NOUN
dpc-1863	49	10	as	as	ADP
dpc-1863	49	11	a	a	DET
dpc-1863	49	12	whole	whole	ADJ
dpc-1863	49	13	,	,	PUNCT
dpc-1863	49	14	object	object	NOUN
dpc-1863	49	15	detection	detection	NOUN
dpc-1863	49	16	(	(	PUNCT
dpc-1863	49	17	ie	ie	X
dpc-1863	49	18	faster	fast	ADJ
dpc-1863	49	19	r	r	NOUN
dpc-1863	49	20	-	-	PUNCT
dpc-1863	49	21	cnn	cnn	NOUN
dpc-1863	49	22	)	)	PUNCT
dpc-1863	49	23	finds	find	VERB
dpc-1863	49	24	objects	object	NOUN
dpc-1863	49	25	and	and	CCONJ
dpc-1863	49	26	their	their	PRON
dpc-1863	49	27	approximate	approximate	ADJ
dpc-1863	49	28	position	position	NOUN
dpc-1863	49	29	in	in	ADP
dpc-1863	49	30	the	the	DET
dpc-1863	49	31	image	image	NOUN
dpc-1863	49	32	and	and	CCONJ
dpc-1863	49	33	instance	instance	NOUN
dpc-1863	49	34	segmentation	segmentation	NOUN
dpc-1863	49	35	(	(	PUNCT
dpc-1863	49	36	ie	ie	X
dpc-1863	49	37	mask	mask	VERB
dpc-1863	49	38	r	r	NOUN
dpc-1863	49	39	-	-	PUNCT
dpc-1863	49	40	cnn	cnn	PROPN
dpc-1863	49	41	)	)	PUNCT
dpc-1863	49	42	finds	find	VERB
dpc-1863	49	43	objects	object	NOUN
dpc-1863	49	44	and	and	CCONJ
dpc-1863	49	45	their	their	PRON
dpc-1863	49	46	exact	exact	ADJ
dpc-1863	49	47	spatial	spatial	ADJ
dpc-1863	49	48	delimitation	delimitation	NOUN
dpc-1863	49	49	.	.	PUNCT
dpc-1863	50	1	original	original	ADJ
dpc-1863	50	2	article	article	NOUN
dpc-1863	50	3	|	|	ADV
dpc-1863	50	4	dermatol	dermatol	ADJ
dpc-1863	50	5	pract	pract	NOUN
dpc-1863	50	6	concept	concept	NOUN
dpc-1863	50	7	.	.	PUNCT
dpc-1863	51	1	2022;12(3):e2022126	2022;12(3):e2022126	NUM
dpc-1863	51	2	3	3	NUM
dpc-1863	51	3	(	(	PUNCT
dpc-1863	51	4	akiec	akiec	NOUN
dpc-1863	51	5	)	)	PUNCT
dpc-1863	51	6	,	,	PUNCT
dpc-1863	51	7	basal	basal	ADJ
dpc-1863	51	8	cell	cell	NOUN
dpc-1863	51	9	carcinoma	carcinoma	NOUN
dpc-1863	51	10	(	(	PUNCT
dpc-1863	51	11	bcc	bcc	PROPN
dpc-1863	51	12	)	)	PUNCT
dpc-1863	51	13	,	,	PUNCT
dpc-1863	51	14	benign	benign	ADJ
dpc-1863	51	15	keratotic	keratotic	ADJ
dpc-1863	51	16	lesion	lesion	NOUN
dpc-1863	51	17	(	(	PUNCT
dpc-1863	51	18	bkl	bkl	NOUN
dpc-1863	51	19	)	)	PUNCT
dpc-1863	51	20	,	,	PUNCT
dpc-1863	51	21	dermatofibroma	dermatofibroma	PROPN
dpc-1863	51	22	(	(	PUNCT
dpc-1863	51	23	df	df	PROPN
dpc-1863	51	24	)	)	PUNCT
dpc-1863	51	25	,	,	PUNCT
dpc-1863	51	26	nevus	nevus	X
dpc-1863	51	27	(	(	PUNCT
dpc-1863	51	28	nv	nv	PROPN
dpc-1863	51	29	)	)	PUNCT
dpc-1863	51	30	,	,	PUNCT
dpc-1863	51	31	melanoma	melanoma	PROPN
dpc-1863	51	32	(	(	PUNCT
dpc-1863	51	33	mel	mel	PROPN
dpc-1863	51	34	)	)	PUNCT
dpc-1863	51	35	,	,	PUNCT
dpc-1863	51	36	or	or	CCONJ
dpc-1863	51	37	vascular	vascular	ADJ
dpc-1863	51	38	lesion	lesion	NOUN
dpc-1863	51	39	(	(	PUNCT
dpc-1863	51	40	vasc	vasc	ADJ
dpc-1863	51	41	)	)	PUNCT
dpc-1863	51	42	.	.	PUNCT
dpc-1863	52	1	also	also	ADV
dpc-1863	52	2	,	,	PUNCT
dpc-1863	52	3	the	the	DET
dpc-1863	52	4	ph2	ph2	NOUN
dpc-1863	52	5	and	and	CCONJ
dpc-1863	52	6	the	the	DET
dpc-1863	52	7	7	7	NUM
dpc-1863	52	8	-	-	PUNCT
dpc-1863	52	9	point	point	NOUN
dpc-1863	52	10	criteria	criterion	NOUN
dpc-1863	52	11	evaluation	evaluation	NOUN
dpc-1863	52	12	dataset	dataset	NOUN
dpc-1863	52	13	were	be	AUX
dpc-1863	52	14	reviewed	review	VERB
dpc-1863	52	15	and	and	CCONJ
dpc-1863	52	16	several	several	ADJ
dpc-1863	52	17	images	image	NOUN
dpc-1863	52	18	were	be	AUX
dpc-1863	52	19	utilized	utilize	VERB
dpc-1863	52	20	to	to	PART
dpc-1863	52	21	extract	extract	VERB
dpc-1863	52	22	artefacts	artefact	NOUN
dpc-1863	52	23	from	from	ADP
dpc-1863	52	24	[	[	X
dpc-1863	52	25	20,21	20,21	NUM
dpc-1863	52	26	]	]	PUNCT
dpc-1863	52	27	.	.	PUNCT
dpc-1863	53	1	images	image	NOUN
dpc-1863	53	2	from	from	ADP
dpc-1863	53	3	those	those	DET
dpc-1863	53	4	datasets	dataset	NOUN
dpc-1863	53	5	were	be	AUX
dpc-1863	53	6	not	not	PART
dpc-1863	53	7	used	use	VERB
dpc-1863	53	8	for	for	ADP
dpc-1863	53	9	other	other	ADJ
dpc-1863	53	10	purposes	purpose	NOUN
dpc-1863	53	11	within	within	ADP
dpc-1863	53	12	this	this	DET
dpc-1863	53	13	study	study	NOUN
dpc-1863	53	14	.	.	PUNCT
dpc-1863	54	1	we	we	PRON
dpc-1863	54	2	used	use	VERB
dpc-1863	54	3	the	the	DET
dpc-1863	54	4	isic2018	isic2018	PROPN
dpc-1863	54	5	test	test	NOUN
dpc-1863	54	6	-	-	PUNCT
dpc-1863	54	7	set	set	NOUN
dpc-1863	54	8	as	as	ADP
dpc-1863	54	9	the	the	DET
dpc-1863	54	10	test	test	NOUN
dpc-1863	54	11	-	-	PUNCT
dpc-1863	54	12	set	set	NOUN
dpc-1863	54	13	to	to	PART
dpc-1863	54	14	keep	keep	VERB
dpc-1863	54	15	variation	variation	NOUN
dpc-1863	54	16	as	as	ADV
dpc-1863	54	17	low	low	ADJ
dpc-1863	54	18	as	as	ADP
dpc-1863	54	19	possible	possible	ADJ
dpc-1863	54	20	,	,	PUNCT
dpc-1863	54	21	as	as	SCONJ
dpc-1863	54	22	it	it	PRON
dpc-1863	54	23	sources	source	VERB
dpc-1863	54	24	from	from	ADP
dpc-1863	54	25	the	the	DET
dpc-1863	54	26	same	same	ADJ
dpc-1863	54	27	origin	origin	NOUN
dpc-1863	54	28	as	as	ADP
dpc-1863	54	29	the	the	DET
dpc-1863	54	30	ham10000	ham10000	PROPN
dpc-1863	54	31	dataset	dataset	NOUN
dpc-1863	54	32	and	and	CCONJ
dpc-1863	54	33	includes	include	VERB
dpc-1863	54	34	the	the	DET
dpc-1863	54	35	same	same	ADJ
dpc-1863	54	36	classes	class	NOUN
dpc-1863	54	37	.	.	PUNCT
dpc-1863	55	1	artefact	artefact	ADJ
dpc-1863	55	2	generation	generation	NOUN
dpc-1863	55	3	as	as	ADP
dpc-1863	55	4	with	with	ADP
dpc-1863	55	5	every	every	DET
dpc-1863	55	6	real	real	ADJ
dpc-1863	55	7	-	-	PUNCT
dpc-1863	55	8	world	world	NOUN
dpc-1863	55	9	picture	picture	NOUN
dpc-1863	55	10	,	,	PUNCT
dpc-1863	55	11	dermatoscopic	dermatoscopic	NOUN
dpc-1863	55	12	images	image	NOUN
dpc-1863	55	13	can	can	AUX
dpc-1863	55	14	contain	contain	VERB
dpc-1863	55	15	content	content	NOUN
dpc-1863	55	16	considered	consider	VERB
dpc-1863	55	17	as	as	ADP
dpc-1863	55	18	“	"	PUNCT
dpc-1863	55	19	artefacts	artefact	NOUN
dpc-1863	55	20	”	"	PUNCT
dpc-1863	55	21	.	.	PUNCT
dpc-1863	56	1	examples	example	NOUN
dpc-1863	56	2	are	be	AUX
dpc-1863	56	3	hairs	hair	NOUN
dpc-1863	56	4	,	,	PUNCT
dpc-1863	56	5	dark	dark	ADJ
dpc-1863	56	6	corners	corner	NOUN
dpc-1863	56	7	,	,	PUNCT
dpc-1863	56	8	vignettes	vignette	NOUN
dpc-1863	56	9	,	,	PUNCT
dpc-1863	56	10	medical	medical	ADJ
dpc-1863	56	11	devices	device	NOUN
dpc-1863	56	12	,	,	PUNCT
dpc-1863	56	13	different	different	ADJ
dpc-1863	56	14	sorts	sort	NOUN
dpc-1863	56	15	of	of	ADP
dpc-1863	56	16	rulers	ruler	NOUN
dpc-1863	56	17	,	,	PUNCT
dpc-1863	56	18	ink	ink	NOUN
dpc-1863	56	19	markings	marking	NOUN
dpc-1863	56	20	in	in	ADP
dpc-1863	56	21	different	different	ADJ
dpc-1863	56	22	shapes	shape	NOUN
dpc-1863	56	23	,	,	PUNCT
dpc-1863	56	24	styles	style	NOUN
dpc-1863	56	25	and	and	CCONJ
dpc-1863	56	26	colors	color	NOUN
dpc-1863	56	27	,	,	PUNCT
dpc-1863	56	28	air	air	NOUN
dpc-1863	56	29	bubbles	bubble	NOUN
dpc-1863	56	30	or	or	CCONJ
dpc-1863	56	31	reflections	reflection	NOUN
dpc-1863	56	32	.	.	PUNCT
dpc-1863	57	1	this	this	DET
dpc-1863	57	2	work	work	NOUN
dpc-1863	57	3	focuses	focus	VERB
dpc-1863	57	4	on	on	ADP
dpc-1863	57	5	three	three	NUM
dpc-1863	57	6	of	of	ADP
dpc-1863	57	7	them	they	PRON
dpc-1863	57	8	:	:	PUNCT
dpc-1863	57	9	“	"	PUNCT
dpc-1863	57	10	bubbles	bubble	NOUN
dpc-1863	57	11	”	"	PUNCT
dpc-1863	57	12	that	that	PRON
dpc-1863	57	13	originate	originate	VERB
dpc-1863	57	14	from	from	ADP
dpc-1863	57	15	trapped	trap	VERB
dpc-1863	57	16	air	air	NOUN
dpc-1863	57	17	in	in	ADP
dpc-1863	57	18	the	the	DET
dpc-1863	57	19	liquid	liquid	NOUN
dpc-1863	57	20	between	between	ADP
dpc-1863	57	21	skin	skin	NOUN
dpc-1863	57	22	and	and	CCONJ
dpc-1863	57	23	the	the	DET
dpc-1863	57	24	dermatoscope	dermatoscope	NOUN
dpc-1863	57	25	,	,	PUNCT
dpc-1863	57	26	“	"	PUNCT
dpc-1863	57	27	rulers	ruler	NOUN
dpc-1863	57	28	”	"	PUNCT
dpc-1863	57	29	used	use	VERB
dpc-1863	57	30	to	to	PART
dpc-1863	57	31	show	show	VERB
dpc-1863	57	32	the	the	DET
dpc-1863	57	33	spatial	spatial	ADJ
dpc-1863	57	34	dimension	dimension	NOUN
dpc-1863	57	35	of	of	ADP
dpc-1863	57	36	a	a	DET
dpc-1863	57	37	lesion	lesion	NOUN
dpc-1863	57	38	,	,	PUNCT
dpc-1863	57	39	and	and	CCONJ
dpc-1863	57	40	ink	ink	NOUN
dpc-1863	57	41	“	"	PUNCT
dpc-1863	57	42	markings	marking	NOUN
dpc-1863	57	43	”	"	PUNCT
dpc-1863	57	44	on	on	ADP
dpc-1863	57	45	the	the	DET
dpc-1863	57	46	patient	patient	NOUN
dpc-1863	57	47	’s	’s	PART
dpc-1863	57	48	skin	skin	NOUN
dpc-1863	57	49	used	use	VERB
dpc-1863	57	50	to	to	PART
dpc-1863	57	51	highlight	highlight	VERB
dpc-1863	57	52	the	the	DET
dpc-1863	57	53	lesion	lesion	NOUN
dpc-1863	57	54	for	for	ADP
dpc-1863	57	55	excision	excision	NOUN
dpc-1863	57	56	or	or	CCONJ
dpc-1863	57	57	review	review	NOUN
dpc-1863	57	58	.	.	PUNCT
dpc-1863	58	1	in	in	ADP
dpc-1863	58	2	order	order	NOUN
dpc-1863	58	3	to	to	PART
dpc-1863	58	4	generate	generate	VERB
dpc-1863	58	5	artefact	artefact	ADJ
dpc-1863	58	6	-	-	PUNCT
dpc-1863	58	7	modified	modify	VERB
dpc-1863	58	8	cases	case	NOUN
dpc-1863	58	9	,	,	PUNCT
dpc-1863	58	10	we	we	PRON
dpc-1863	58	11	selected	select	VERB
dpc-1863	58	12	60	60	NUM
dpc-1863	58	13	images	image	NOUN
dpc-1863	58	14	from	from	ADP
dpc-1863	58	15	the	the	DET
dpc-1863	58	16	ham10000	ham10000	NOUN
dpc-1863	58	17	,	,	PUNCT
dpc-1863	58	18	ph2	ph2	NOUN
dpc-1863	58	19	and	and	CCONJ
dpc-1863	58	20	7	7	NUM
dpc-1863	58	21	-	-	PUNCT
dpc-1863	58	22	point	point	NOUN
dpc-1863	58	23	criteria	criterion	NOUN
dpc-1863	58	24	dataset	dataset	VERB
dpc-1863	58	25	which	which	PRON
dpc-1863	58	26	contain	contain	VERB
dpc-1863	58	27	either	either	CCONJ
dpc-1863	58	28	a	a	DET
dpc-1863	58	29	bubble	bubble	NOUN
dpc-1863	58	30	,	,	PUNCT
dpc-1863	58	31	a	a	DET
dpc-1863	58	32	ruler	ruler	NOUN
dpc-1863	58	33	or	or	CCONJ
dpc-1863	58	34	a	a	DET
dpc-1863	58	35	marking	mark	VERB
dpc-1863	58	36	artefact	artefact	NOUN
dpc-1863	58	37	.	.	PUNCT
dpc-1863	59	1	from	from	ADP
dpc-1863	59	2	those	those	DET
dpc-1863	59	3	images	image	NOUN
dpc-1863	59	4	we	we	PRON
dpc-1863	59	5	extracted	extract	VERB
dpc-1863	59	6	the	the	DET
dpc-1863	59	7	artefacts	artefact	NOUN
dpc-1863	59	8	by	by	ADP
dpc-1863	59	9	manually	manually	ADV
dpc-1863	59	10	repairing	repair	VERB
dpc-1863	59	11	the	the	DET
dpc-1863	59	12	images	image	NOUN
dpc-1863	59	13	areas	area	NOUN
dpc-1863	59	14	with	with	ADP
dpc-1863	59	15	adobe	adobe	PROPN
dpc-1863	59	16	®	®	PROPN
dpc-1863	59	17	photoshop	photoshop	PROPN
dpc-1863	59	18	’s	’s	PART
dpc-1863	59	19	®	®	NOUN
dpc-1863	59	20	(	(	PUNCT
dpc-1863	59	21	version	version	NOUN
dpc-1863	59	22	cc	cc	PROPN
dpc-1863	59	23	2018	2018	NUM
dpc-1863	59	24	(	(	PUNCT
dpc-1863	59	25	19.1.9	19.1.9	NUM
dpc-1863	59	26	)	)	PUNCT
dpc-1863	59	27	,	,	PUNCT
dpc-1863	59	28	adobe	adobe	PROPN
dpc-1863	59	29	inc	inc	PROPN
dpc-1863	59	30	.	.	PROPN
dpc-1863	59	31	)	)	PUNCT
dpc-1863	59	32	content	content	ADJ
dpc-1863	59	33	aware	aware	ADJ
dpc-1863	59	34	image	image	NOUN
dpc-1863	59	35	repair	repair	NOUN
dpc-1863	59	36	mechanism	mechanism	NOUN
dpc-1863	59	37	and	and	CCONJ
dpc-1863	59	38	using	use	VERB
dpc-1863	59	39	the	the	DET
dpc-1863	59	40	difference	difference	NOUN
dpc-1863	59	41	,	,	PUNCT
dpc-1863	59	42	per	per	ADP
dpc-1863	59	43	rgb	rgb	PROPN
dpc-1863	59	44	channel	channel	PROPN
dpc-1863	59	45	,	,	PUNCT
dpc-1863	59	46	to	to	ADP
dpc-1863	59	47	the	the	DET
dpc-1863	59	48	untouched	untouched	ADJ
dpc-1863	59	49	image	image	NOUN
dpc-1863	59	50	as	as	ADP
dpc-1863	59	51	a	a	DET
dpc-1863	59	52	template	template	NOUN
dpc-1863	59	53	(	(	PUNCT
dpc-1863	59	54	figure	figure	NOUN
dpc-1863	59	55	2	2	NUM
dpc-1863	59	56	)	)	PUNCT
dpc-1863	59	57	.	.	PUNCT
dpc-1863	60	1	the	the	DET
dpc-1863	60	2	insertion	insertion	NOUN
dpc-1863	60	3	of	of	ADP
dpc-1863	60	4	those	those	DET
dpc-1863	60	5	templates	template	NOUN
dpc-1863	60	6	was	be	AUX
dpc-1863	60	7	done	do	VERB
dpc-1863	60	8	in	in	ADP
dpc-1863	60	9	a	a	DET
dpc-1863	60	10	way	way	NOUN
dpc-1863	60	11	that	that	PRON
dpc-1863	60	12	the	the	DET
dpc-1863	60	13	position	position	NOUN
dpc-1863	60	14	of	of	ADP
dpc-1863	60	15	artefacts	artefact	NOUN
dpc-1863	60	16	varies	vary	VERB
dpc-1863	60	17	according	accord	VERB
dpc-1863	60	18	to	to	ADP
dpc-1863	60	19	observed	observe	VERB
dpc-1863	60	20	patterns	pattern	NOUN
dpc-1863	60	21	,	,	PUNCT
dpc-1863	60	22	using	use	VERB
dpc-1863	60	23	the	the	DET
dpc-1863	60	24	provided	provide	VERB
dpc-1863	60	25	segmentation	segmentation	NOUN
dpc-1863	60	26	mask	mask	NOUN
dpc-1863	60	27	of	of	ADP
dpc-1863	60	28	the	the	DET
dpc-1863	60	29	target	target	NOUN
dpc-1863	60	30	image	image	NOUN
dpc-1863	60	31	.	.	PUNCT
dpc-1863	61	1	in	in	ADP
dpc-1863	61	2	figure	figure	NOUN
dpc-1863	61	3	3	3	NUM
dpc-1863	61	4	,	,	PUNCT
dpc-1863	61	5	a	a	DET
dpc-1863	61	6	dermatoscopic	dermatoscopic	ADJ
dpc-1863	61	7	image	image	NOUN
dpc-1863	61	8	with	with	ADP
dpc-1863	61	9	automatically	automatically	ADV
dpc-1863	61	10	superimposed	superimpose	VERB
dpc-1863	61	11	artefacts	artefact	NOUN
dpc-1863	61	12	is	be	AUX
dpc-1863	61	13	shown	show	VERB
dpc-1863	61	14	.	.	PUNCT
dpc-1863	62	1	the	the	DET
dpc-1863	62	2	source	source	NOUN
dpc-1863	62	3	code	code	NOUN
dpc-1863	62	4	will	will	AUX
dpc-1863	62	5	be	be	AUX
dpc-1863	62	6	made	make	VERB
dpc-1863	62	7	available	available	ADJ
dpc-1863	62	8	upon	upon	SCONJ
dpc-1863	62	9	publication	publication	NOUN
dpc-1863	62	10	of	of	ADP
dpc-1863	62	11	this	this	DET
dpc-1863	62	12	work	work	NOUN
dpc-1863	62	13	at	at	ADP
dpc-1863	62	14	https://github.com/thisismexp/artefact_insertion	https://github.com/thisismexp/artefact_insertion	NOUN
dpc-1863	62	15	.	.	PUNCT
dpc-1863	63	1	using	use	VERB
dpc-1863	63	2	the	the	DET
dpc-1863	63	3	artefact	artefact	ADJ
dpc-1863	63	4	insertion	insertion	NOUN
dpc-1863	63	5	mechanism	mechanism	NOUN
dpc-1863	63	6	,	,	PUNCT
dpc-1863	63	7	several	several	ADJ
dpc-1863	63	8	dataset	dataset	ADJ
dpc-1863	63	9	mutations	mutation	NOUN
dpc-1863	63	10	of	of	ADP
dpc-1863	63	11	the	the	DET
dpc-1863	63	12	original	original	ADJ
dpc-1863	63	13	ham10000	ham10000	NOUN
dpc-1863	63	14	dataset	dataset	NOUN
dpc-1863	63	15	were	be	AUX
dpc-1863	63	16	created	create	VERB
dpc-1863	63	17	,	,	PUNCT
dpc-1863	63	18	where	where	SCONJ
dpc-1863	63	19	artefacts	artefact	NOUN
dpc-1863	63	20	were	be	AUX
dpc-1863	63	21	superimposed	superimpose	VERB
dpc-1863	63	22	on	on	ADP
dpc-1863	63	23	either	either	CCONJ
dpc-1863	63	24	none	none	NOUN
dpc-1863	63	25	or	or	CCONJ
dpc-1863	63	26	all	all	PRON
dpc-1863	63	27	of	of	ADP
dpc-1863	63	28	the	the	DET
dpc-1863	63	29	images	image	NOUN
dpc-1863	63	30	and	and	CCONJ
dpc-1863	63	31	on	on	ADP
dpc-1863	63	32	every	every	DET
dpc-1863	63	33	image	image	NOUN
dpc-1863	63	34	belonging	belong	VERB
dpc-1863	63	35	to	to	ADP
dpc-1863	63	36	a	a	DET
dpc-1863	63	37	certain	certain	ADJ
dpc-1863	63	38	diagnosis	diagnosis	NOUN
dpc-1863	63	39	.	.	PUNCT
dpc-1863	64	1	the	the	DET
dpc-1863	64	2	test	test	NOUN
dpc-1863	64	3	portion	portion	NOUN
dpc-1863	64	4	of	of	ADP
dpc-1863	64	5	the	the	DET
dpc-1863	64	6	ham10000	ham10000	PROPN
dpc-1863	64	7	dataset	dataset	PROPN
dpc-1863	64	8	,	,	PUNCT
dpc-1863	64	9	corresponding	correspond	VERB
dpc-1863	64	10	to	to	ADP
dpc-1863	64	11	the	the	DET
dpc-1863	64	12	isic2018	isic2018	PROPN
dpc-1863	64	13	challenge	challenge	PROPN
dpc-1863	64	14	task	task	PROPN
dpc-1863	64	15	3	3	NUM
dpc-1863	64	16	test	test	NOUN
dpc-1863	64	17	-	-	PUNCT
dpc-1863	64	18	set	set	VERB
dpc-1863	64	19	with	with	ADP
dpc-1863	64	20	1,511	1,511	NUM
dpc-1863	64	21	images	image	NOUN
dpc-1863	64	22	,	,	PUNCT
dpc-1863	64	23	was	be	AUX
dpc-1863	64	24	altered	alter	VERB
dpc-1863	64	25	in	in	ADP
dpc-1863	64	26	the	the	DET
dpc-1863	64	27	same	same	ADJ
dpc-1863	64	28	way	way	NOUN
dpc-1863	64	29	.	.	PUNCT
dpc-1863	65	1	additionally	additionally	ADV
dpc-1863	65	2	,	,	PUNCT
dpc-1863	65	3	artefacts	artefact	NOUN
dpc-1863	65	4	were	be	AUX
dpc-1863	65	5	inserted	insert	VERB
dpc-1863	65	6	in	in	ADP
dpc-1863	65	7	a	a	DET
dpc-1863	65	8	certain	certain	ADJ
dpc-1863	65	9	percentage	percentage	NOUN
dpc-1863	65	10	of	of	ADP
dpc-1863	65	11	images	image	NOUN
dpc-1863	65	12	in	in	ADP
dpc-1863	65	13	20	20	NUM
dpc-1863	65	14	%	%	NOUN
dpc-1863	65	15	step	step	NOUN
dpc-1863	65	16	increments	increment	NOUN
dpc-1863	65	17	.	.	PUNCT
dpc-1863	66	1	neural	neural	ADJ
dpc-1863	66	2	network	network	NOUN
dpc-1863	66	3	training	training	NOUN
dpc-1863	66	4	as	as	ADP
dpc-1863	66	5	representatives	representative	NOUN
dpc-1863	66	6	for	for	ADP
dpc-1863	66	7	image	image	NOUN
dpc-1863	66	8	classification	classification	NOUN
dpc-1863	66	9	,	,	PUNCT
dpc-1863	66	10	object	object	NOUN
dpc-1863	66	11	detection	detection	NOUN
dpc-1863	66	12	and	and	CCONJ
dpc-1863	66	13	instance	instance	NOUN
dpc-1863	66	14	segmentation	segmentation	NOUN
dpc-1863	66	15	we	we	PRON
dpc-1863	66	16	trained	train	VERB
dpc-1863	66	17	a	a	DET
dpc-1863	66	18	resnet-34	resnet-34	PROPN
dpc-1863	66	19	,	,	PUNCT
dpc-1863	66	20	a	a	DET
dpc-1863	66	21	faster	fast	ADJ
dpc-1863	66	22	input	input	NOUN
dpc-1863	66	23	:	:	PUNCT
dpc-1863	66	24	original	original	ADJ
dpc-1863	66	25	image	image	NOUN
dpc-1863	66	26	output	output	NOUN
dpc-1863	66	27	:	:	PUNCT
dpc-1863	66	28	artefact	artefact	ADJ
dpc-1863	66	29	template	template	NOUN
dpc-1863	66	30	1	1	NUM
dpc-1863	66	31	.	.	PUNCT
dpc-1863	66	32	)	)	PUNCT
dpc-1863	67	1	content	content	NOUN
dpc-1863	67	2	-	-	PUNCT
dpc-1863	67	3	aware	aware	ADJ
dpc-1863	67	4	image	image	NOUN
dpc-1863	67	5	repair	repair	NOUN
dpc-1863	67	6	3	3	NUM
dpc-1863	67	7	.	.	PUNCT
dpc-1863	67	8	)	)	PUNCT
dpc-1863	68	1	rgb	rgb	PROPN
dpc-1863	68	2	channel	channel	PROPN
dpc-1863	68	3	-	-	PUNCT
dpc-1863	68	4	wise	wise	ADJ
dpc-1863	68	5	di	di	NOUN
dpc-1863	68	6	�	�	PROPN
dpc-1863	68	7	erence	erence	NOUN
dpc-1863	68	8	2	2	NUM
dpc-1863	68	9	.	.	PUNCT
dpc-1863	68	10	)	)	PUNCT
dpc-1863	69	1	extraction	extraction	NOUN
dpc-1863	69	2	of	of	ADP
dpc-1863	69	3	repaired	repair	VERB
dpc-1863	69	4	regions	region	NOUN
dpc-1863	69	5	figure	figure	VERB
dpc-1863	69	6	2	2	NUM
dpc-1863	69	7	.	.	PUNCT
dpc-1863	69	8	workflow	workflow	NOUN
dpc-1863	69	9	for	for	ADP
dpc-1863	69	10	extracting	extract	VERB
dpc-1863	69	11	artefact	artefact	ADJ
dpc-1863	69	12	templates	template	NOUN
dpc-1863	69	13	.	.	PUNCT
dpc-1863	70	1	manually	manually	ADV
dpc-1863	70	2	selected	select	VERB
dpc-1863	70	3	original	original	ADJ
dpc-1863	70	4	images	image	NOUN
dpc-1863	70	5	(	(	PUNCT
dpc-1863	70	6	input	input	NOUN
dpc-1863	70	7	)	)	PUNCT
dpc-1863	70	8	were	be	AUX
dpc-1863	70	9	repaired	repair	VERB
dpc-1863	70	10	manually	manually	ADV
dpc-1863	70	11	(	(	PUNCT
dpc-1863	70	12	1	1	NUM
dpc-1863	70	13	)	)	PUNCT
dpc-1863	70	14	,	,	PUNCT
dpc-1863	70	15	and	and	CCONJ
dpc-1863	70	16	corresponding	correspond	VERB
dpc-1863	70	17	image	image	NOUN
dpc-1863	70	18	areas	area	NOUN
dpc-1863	70	19	extracted	extract	VERB
dpc-1863	70	20	(	(	PUNCT
dpc-1863	70	21	2	2	NUM
dpc-1863	70	22	)	)	PUNCT
dpc-1863	70	23	.	.	PUNCT
dpc-1863	71	1	the	the	DET
dpc-1863	71	2	channel	channel	NOUN
dpc-1863	71	3	-	-	PUNCT
dpc-1863	71	4	wise	wise	ADJ
dpc-1863	71	5	difference	difference	NOUN
dpc-1863	71	6	(	(	PUNCT
dpc-1863	71	7	3	3	X
dpc-1863	71	8	)	)	PUNCT
dpc-1863	71	9	was	be	AUX
dpc-1863	71	10	stored	store	VERB
dpc-1863	71	11	as	as	ADP
dpc-1863	71	12	a	a	DET
dpc-1863	71	13	template	template	NOUN
dpc-1863	71	14	for	for	ADP
dpc-1863	71	15	the	the	DET
dpc-1863	71	16	corresponding	corresponding	ADJ
dpc-1863	71	17	artefact	artefact	ADJ
dpc-1863	71	18	type	type	NOUN
dpc-1863	71	19	.	.	PUNCT
dpc-1863	72	1	4	4	NUM
dpc-1863	72	2	original	original	ADJ
dpc-1863	72	3	article	article	NOUN
dpc-1863	72	4	|	|	ADV
dpc-1863	72	5	dermatol	dermatol	ADJ
dpc-1863	72	6	pract	pract	NOUN
dpc-1863	72	7	concept	concept	NOUN
dpc-1863	72	8	.	.	PUNCT
dpc-1863	73	1	2022;12(3):e2022126	2022;12(3):e2022126	NUM
dpc-1863	73	2	r	r	X
dpc-1863	73	3	-	-	PUNCT
dpc-1863	73	4	cnn	cnn	PROPN
dpc-1863	73	5	(	(	PUNCT
dpc-1863	73	6	with	with	ADP
dpc-1863	73	7	resnet-34	resnet-34	PROPN
dpc-1863	73	8	backbone	backbone	NOUN
dpc-1863	73	9	)	)	PUNCT
dpc-1863	73	10	and	and	CCONJ
dpc-1863	73	11	a	a	DET
dpc-1863	73	12	mask	mask	NOUN
dpc-1863	73	13	r	r	NOUN
dpc-1863	73	14	-	-	PUNCT
dpc-1863	73	15	cnn	cnn	PROPN
dpc-1863	73	16	(	(	PUNCT
dpc-1863	73	17	also	also	ADV
dpc-1863	73	18	with	with	ADP
dpc-1863	73	19	a	a	DET
dpc-1863	73	20	resnet-34	resnet-34	PROPN
dpc-1863	73	21	backbone	backbone	NOUN
dpc-1863	73	22	)	)	PUNCT
dpc-1863	73	23	model	model	NOUN
dpc-1863	73	24	as	as	SCONJ
dpc-1863	73	25	provided	provide	VERB
dpc-1863	73	26	by	by	ADP
dpc-1863	73	27	the	the	DET
dpc-1863	73	28	torchvision	torchvision	NOUN
dpc-1863	73	29	package	package	NOUN
dpc-1863	73	30	of	of	ADP
dpc-1863	73	31	the	the	DET
dpc-1863	73	32	open	open	ADJ
dpc-1863	73	33	source	source	NOUN
dpc-1863	73	34	machine	machine	NOUN
dpc-1863	73	35	learning	learn	VERB
dpc-1863	73	36	framework	framework	NOUN
dpc-1863	73	37	pytorch	pytorch	NOUN
dpc-1863	74	1	[	[	X
dpc-1863	74	2	22	22	NUM
dpc-1863	74	3	]	]	PUNCT
dpc-1863	74	4	.	.	PUNCT
dpc-1863	75	1	all	all	DET
dpc-1863	75	2	models	model	NOUN
dpc-1863	75	3	were	be	AUX
dpc-1863	75	4	trained	train	VERB
dpc-1863	75	5	on	on	ADP
dpc-1863	75	6	all	all	PRON
dpc-1863	75	7	of	of	ADP
dpc-1863	75	8	the	the	DET
dpc-1863	75	9	9	9	NUM
dpc-1863	75	10	generated	generate	VERB
dpc-1863	75	11	datasets	dataset	NOUN
dpc-1863	75	12	in	in	ADP
dpc-1863	75	13	a	a	DET
dpc-1863	75	14	5	5	NUM
dpc-1863	75	15	-	-	ADJ
dpc-1863	75	16	fold	fold	ADJ
dpc-1863	75	17	cross	cross	NOUN
dpc-1863	75	18	validation	validation	NOUN
dpc-1863	75	19	fashion	fashion	NOUN
dpc-1863	75	20	.	.	PUNCT
dpc-1863	76	1	transfer	transfer	NOUN
dpc-1863	76	2	-	-	PUNCT
dpc-1863	76	3	learning	learn	VERB
dpc-1863	76	4	and	and	CCONJ
dpc-1863	76	5	data	datum	NOUN
dpc-1863	76	6	augmentation	augmentation	NOUN
dpc-1863	76	7	including	include	VERB
dpc-1863	76	8	random	random	ADJ
dpc-1863	76	9	crops	crop	NOUN
dpc-1863	76	10	,	,	PUNCT
dpc-1863	76	11	resize	resize	NOUN
dpc-1863	76	12	,	,	PUNCT
dpc-1863	76	13	rotations	rotation	NOUN
dpc-1863	76	14	,	,	PUNCT
dpc-1863	76	15	mirroring	mirror	VERB
dpc-1863	76	16	operations	operation	NOUN
dpc-1863	76	17	as	as	ADV
dpc-1863	76	18	well	well	ADV
dpc-1863	76	19	as	as	ADP
dpc-1863	76	20	color	color	NOUN
dpc-1863	76	21	jitter	jitter	NOUN
dpc-1863	76	22	operations	operation	NOUN
dpc-1863	76	23	were	be	AUX
dpc-1863	76	24	used	use	VERB
dpc-1863	76	25	.	.	PUNCT
dpc-1863	77	1	statistics	statistic	NOUN
dpc-1863	77	2	to	to	PART
dpc-1863	77	3	evaluate	evaluate	VERB
dpc-1863	77	4	diagnostic	diagnostic	ADJ
dpc-1863	77	5	accuracy	accuracy	NOUN
dpc-1863	77	6	,	,	PUNCT
dpc-1863	77	7	all	all	DET
dpc-1863	77	8	trained	train	VERB
dpc-1863	77	9	network	network	NOUN
dpc-1863	77	10	models	model	NOUN
dpc-1863	77	11	are	be	AUX
dpc-1863	77	12	tested	test	VERB
dpc-1863	77	13	against	against	ADP
dpc-1863	77	14	the	the	DET
dpc-1863	77	15	13	13	NUM
dpc-1863	77	16	test	test	NOUN
dpc-1863	77	17	datasets	dataset	NOUN
dpc-1863	77	18	and	and	CCONJ
dpc-1863	77	19	performance	performance	NOUN
dpc-1863	77	20	was	be	AUX
dpc-1863	77	21	reported	report	VERB
dpc-1863	77	22	in	in	ADP
dpc-1863	77	23	terms	term	NOUN
dpc-1863	77	24	of	of	ADP
dpc-1863	77	25	area	area	NOUN
dpc-1863	77	26	under	under	ADP
dpc-1863	77	27	the	the	DET
dpc-1863	77	28	precision	precision	NOUN
dpc-1863	77	29	recall	recall	NOUN
dpc-1863	77	30	curve	curve	NOUN
dpc-1863	77	31	(	(	PUNCT
dpc-1863	77	32	pr	pr	NOUN
dpc-1863	77	33	-	-	PUNCT
dpc-1863	77	34	auc	auc	NOUN
dpc-1863	77	35	)	)	PUNCT
dpc-1863	77	36	,	,	PUNCT
dpc-1863	77	37	precision	precision	NOUN
dpc-1863	77	38	,	,	PUNCT
dpc-1863	77	39	recall	recall	NOUN
dpc-1863	77	40	,	,	PUNCT
dpc-1863	77	41	false	false	ADJ
dpc-1863	77	42	positive	positive	ADJ
dpc-1863	77	43	(	(	PUNCT
dpc-1863	77	44	fpr	fpr	NOUN
dpc-1863	77	45	)	)	PUNCT
dpc-1863	77	46	and	and	CCONJ
dpc-1863	77	47	false	false	ADJ
dpc-1863	77	48	negative	negative	ADJ
dpc-1863	77	49	rates	rate	NOUN
dpc-1863	77	50	(	(	PUNCT
dpc-1863	77	51	fnr	fnr	PROPN
dpc-1863	77	52	)	)	PUNCT
dpc-1863	77	53	and	and	CCONJ
dpc-1863	77	54	differences	difference	NOUN
dpc-1863	77	55	thereof	thereof	ADV
dpc-1863	77	56	(	(	PUNCT
dpc-1863	77	57	calculated	calculate	VERB
dpc-1863	77	58	using	use	VERB
dpc-1863	77	59	scikit	scikit	NOUN
dpc-1863	77	60	-	-	PUNCT
dpc-1863	77	61	learn	learn	VERB
dpc-1863	77	62	version	version	NOUN
dpc-1863	77	63	0.24.1	0.24.1	NOUN
dpc-1863	77	64	)	)	PUNCT
dpc-1863	78	1	[	[	X
dpc-1863	78	2	23	23	NUM
dpc-1863	78	3	]	]	PUNCT
dpc-1863	78	4	.	.	PUNCT
dpc-1863	79	1	to	to	PART
dpc-1863	79	2	visualize	visualize	VERB
dpc-1863	79	3	spatial	spatial	ADJ
dpc-1863	79	4	activations	activation	NOUN
dpc-1863	79	5	,	,	PUNCT
dpc-1863	79	6	gradient	gradient	ADJ
dpc-1863	79	7	based	base	VERB
dpc-1863	79	8	class	class	NOUN
dpc-1863	79	9	activation	activation	NOUN
dpc-1863	79	10	map	map	NOUN
dpc-1863	79	11	(	(	PUNCT
dpc-1863	79	12	grad	grad	NOUN
dpc-1863	79	13	-	-	PUNCT
dpc-1863	79	14	cam	cam	NOUN
dpc-1863	79	15	)	)	PUNCT
dpc-1863	79	16	visualizations	visualization	NOUN
dpc-1863	79	17	were	be	AUX
dpc-1863	79	18	used	use	VERB
dpc-1863	79	19	.	.	PUNCT
dpc-1863	80	1	a	a	DET
dpc-1863	80	2	two	two	NUM
dpc-1863	80	3	-	-	PUNCT
dpc-1863	80	4	sided	sided	ADJ
dpc-1863	80	5	p	p	NOUN
dpc-1863	80	6	-	-	PUNCT
dpc-1863	80	7	value	value	NOUN
dpc-1863	80	8	of	of	ADP
dpc-1863	80	9	0.05	0.05	NUM
dpc-1863	80	10	was	be	AUX
dpc-1863	80	11	regarded	regard	VERB
dpc-1863	80	12	as	as	ADP
dpc-1863	80	13	statistically	statistically	ADV
dpc-1863	80	14	significant	significant	ADJ
dpc-1863	80	15	,	,	PUNCT
dpc-1863	80	16	and	and	CCONJ
dpc-1863	80	17	all	all	DET
dpc-1863	80	18	calculations	calculation	NOUN
dpc-1863	80	19	were	be	AUX
dpc-1863	80	20	performed	perform	VERB
dpc-1863	80	21	using	use	VERB
dpc-1863	80	22	statsmodels	statsmodel	NOUN
dpc-1863	80	23	version	version	NOUN
dpc-1863	80	24	0.12.2	0.12.2	PUNCT
dpc-1863	81	1	[	[	X
dpc-1863	81	2	24	24	NUM
dpc-1863	81	3	]	]	PUNCT
dpc-1863	81	4	.	.	PUNCT
dpc-1863	82	1	results	result	NOUN
dpc-1863	82	2	baseline	baseline	VERB
dpc-1863	82	3	performance	performance	NOUN
dpc-1863	82	4	in	in	ADP
dpc-1863	82	5	terms	term	NOUN
dpc-1863	82	6	of	of	ADP
dpc-1863	82	7	pr	pr	NOUN
dpc-1863	82	8	-	-	PUNCT
dpc-1863	82	9	auc	auc	NOUN
dpc-1863	82	10	of	of	ADP
dpc-1863	82	11	our	our	PRON
dpc-1863	82	12	models	model	NOUN
dpc-1863	82	13	trained	train	VERB
dpc-1863	82	14	and	and	CCONJ
dpc-1863	82	15	tested	test	VERB
dpc-1863	82	16	with	with	ADP
dpc-1863	82	17	no	no	DET
dpc-1863	82	18	additional	additional	ADJ
dpc-1863	82	19	artefacts	artefact	NOUN
dpc-1863	82	20	was	be	AUX
dpc-1863	82	21	0.8	0.8	NUM
dpc-1863	82	22	for	for	ADP
dpc-1863	82	23	resnet-34	resnet-34	PROPN
dpc-1863	82	24	and	and	CCONJ
dpc-1863	82	25	0.72	0.72	NUM
dpc-1863	82	26	for	for	ADP
dpc-1863	82	27	faster	fast	ADJ
dpc-1863	82	28	r	r	NOUN
dpc-1863	82	29	-	-	PUNCT
dpc-1863	82	30	cnn	cnn	PROPN
dpc-1863	82	31	as	as	ADV
dpc-1863	82	32	well	well	ADV
dpc-1863	82	33	as	as	ADP
dpc-1863	82	34	mask	mask	NOUN
dpc-1863	82	35	r	r	NOUN
dpc-1863	82	36	-	-	PUNCT
dpc-1863	82	37	cnn	cnn	PROPN
dpc-1863	82	38	.	.	PUNCT
dpc-1863	83	1	introduction	introduction	NOUN
dpc-1863	83	2	of	of	ADP
dpc-1863	83	3	artefacts	artefact	NOUN
dpc-1863	83	4	in	in	ADP
dpc-1863	83	5	only	only	ADV
dpc-1863	83	6	the	the	DET
dpc-1863	83	7	test	test	NOUN
dpc-1863	83	8	dataset	dataset	NOUN
dpc-1863	83	9	led	lead	VERB
dpc-1863	83	10	to	to	ADP
dpc-1863	83	11	a	a	DET
dpc-1863	83	12	reduction	reduction	NOUN
dpc-1863	83	13	in	in	ADP
dpc-1863	83	14	performance	performance	NOUN
dpc-1863	83	15	for	for	ADP
dpc-1863	83	16	all	all	DET
dpc-1863	83	17	three	three	NUM
dpc-1863	83	18	architectures	architecture	NOUN
dpc-1863	83	19	(	(	PUNCT
dpc-1863	83	20	figure	figure	NOUN
dpc-1863	83	21	4	4	NUM
dpc-1863	83	22	)	)	PUNCT
dpc-1863	83	23	increasing	increase	VERB
dpc-1863	83	24	with	with	ADP
dpc-1863	83	25	the	the	DET
dpc-1863	83	26	proportion	proportion	NOUN
dpc-1863	83	27	of	of	ADP
dpc-1863	83	28	artefacts	artefact	NOUN
dpc-1863	83	29	present	present	ADJ
dpc-1863	83	30	in	in	ADP
dpc-1863	83	31	the	the	DET
dpc-1863	83	32	test	test	NOUN
dpc-1863	83	33	dataset	dataset	NOUN
dpc-1863	83	34	,	,	PUNCT
dpc-1863	83	35	and	and	CCONJ
dpc-1863	83	36	more	more	ADV
dpc-1863	83	37	severe	severe	ADJ
dpc-1863	83	38	for	for	ADP
dpc-1863	83	39	the	the	DET
dpc-1863	83	40	resnet-34	resnet-34	PROPN
dpc-1863	83	41	and	and	CCONJ
dpc-1863	83	42	faster	fast	ADJ
dpc-1863	83	43	r	r	NOUN
dpc-1863	83	44	-	-	PUNCT
dpc-1863	83	45	cnn	cnn	PROPN
dpc-1863	83	46	model	model	NOUN
dpc-1863	83	47	.	.	PUNCT
dpc-1863	84	1	with	with	ADP
dpc-1863	84	2	a	a	DET
dpc-1863	84	3	maximum	maximum	ADJ
dpc-1863	84	4	relative	relative	ADJ
dpc-1863	84	5	reduction	reduction	NOUN
dpc-1863	84	6	of	of	ADP
dpc-1863	84	7	0.05	0.05	NUM
dpc-1863	84	8	pr	pr	NOUN
dpc-1863	84	9	-	-	PUNCT
dpc-1863	84	10	auc	auc	NOUN
dpc-1863	84	11	the	the	DET
dpc-1863	84	12	faster	fast	ADJ
dpc-1863	84	13	r	r	NOUN
dpc-1863	84	14	-	-	PUNCT
dpc-1863	84	15	cnn	cnn	PROPN
dpc-1863	84	16	model	model	NOUN
dpc-1863	84	17	was	be	AUX
dpc-1863	84	18	affected	affect	VERB
dpc-1863	84	19	the	the	DET
dpc-1863	84	20	most	most	ADJ
dpc-1863	84	21	,	,	PUNCT
dpc-1863	84	22	resnet-34	resnet-34	PROPN
dpc-1863	84	23	(	(	PUNCT
dpc-1863	84	24	-0.03	-0.03	NOUN
dpc-1863	84	25	)	)	PUNCT
dpc-1863	84	26	the	the	DET
dpc-1863	84	27	second	second	ADJ
dpc-1863	84	28	most	most	ADJ
dpc-1863	84	29	,	,	PUNCT
dpc-1863	84	30	and	and	CCONJ
dpc-1863	84	31	mask	mask	VERB
dpc-1863	84	32	r	r	NOUN
dpc-1863	84	33	-	-	PUNCT
dpc-1863	84	34	cnn	cnn	PROPN
dpc-1863	84	35	was	be	AUX
dpc-1863	84	36	the	the	DET
dpc-1863	84	37	most	most	ADV
dpc-1863	84	38	robust	robust	ADJ
dpc-1863	84	39	(	(	PUNCT
dpc-1863	84	40	-0.01	-0.01	NUM
dpc-1863	84	41	)	)	PUNCT
dpc-1863	84	42	.	.	PUNCT
dpc-1863	85	1	for	for	ADP
dpc-1863	85	2	resnet-34	resnet-34	PROPN
dpc-1863	85	3	and	and	CCONJ
dpc-1863	85	4	faster	fast	ADJ
dpc-1863	85	5	r	r	NOUN
dpc-1863	85	6	-	-	PUNCT
dpc-1863	85	7	cnn	cnn	PROPN
dpc-1863	85	8	,	,	PUNCT
dpc-1863	85	9	changes	change	NOUN
dpc-1863	85	10	in	in	ADP
dpc-1863	85	11	predictive	predictive	ADJ
dpc-1863	85	12	performance	performance	NOUN
dpc-1863	85	13	compared	compare	VERB
dpc-1863	85	14	to	to	PART
dpc-1863	85	15	baseline	baseline	VERB
dpc-1863	85	16	was	be	AUX
dpc-1863	85	17	significant	significant	ADJ
dpc-1863	85	18	at	at	ADP
dpc-1863	85	19	and	and	CCONJ
dpc-1863	85	20	above	above	ADP
dpc-1863	85	21	40	40	NUM
dpc-1863	85	22	%	%	NOUN
dpc-1863	85	23	of	of	ADP
dpc-1863	85	24	introduced	introduce	VERB
dpc-1863	85	25	artefacts	artefact	NOUN
dpc-1863	85	26	in	in	ADP
dpc-1863	85	27	the	the	DET
dpc-1863	85	28	test	test	NOUN
dpc-1863	85	29	set	set	NOUN
dpc-1863	85	30	(	(	PUNCT
dpc-1863	85	31	p	p	X
dpc-1863	85	32	<	<	X
dpc-1863	85	33	0.01	0.01	NUM
dpc-1863	85	34	;	;	PUNCT
dpc-1863	85	35	tested	test	VERB
dpc-1863	85	36	using	use	VERB
dpc-1863	85	37	mcnemar	mcnemar	PROPN
dpc-1863	85	38	test	test	NOUN
dpc-1863	85	39	with	with	ADP
dpc-1863	85	40	edwards	edwards	PROPN
dpc-1863	85	41	correction	correction	NOUN
dpc-1863	85	42	on	on	ADP
dpc-1863	85	43	binarized	binarized	ADJ
dpc-1863	85	44	predictions	prediction	NOUN
dpc-1863	85	45	)	)	PUNCT
dpc-1863	85	46	.	.	PUNCT
dpc-1863	86	1	for	for	ADP
dpc-1863	86	2	mask	mask	NOUN
dpc-1863	86	3	r	r	NOUN
dpc-1863	86	4	-	-	PUNCT
dpc-1863	86	5	cnn	cnn	NOUN
dpc-1863	86	6	we	we	PRON
dpc-1863	86	7	did	do	AUX
dpc-1863	86	8	not	not	PART
dpc-1863	86	9	detect	detect	VERB
dpc-1863	86	10	a	a	DET
dpc-1863	86	11	significant	significant	ADJ
dpc-1863	86	12	difference	difference	NOUN
dpc-1863	86	13	in	in	ADP
dpc-1863	86	14	predictions	prediction	NOUN
dpc-1863	86	15	in	in	ADP
dpc-1863	86	16	all	all	DET
dpc-1863	86	17	used	used	ADJ
dpc-1863	86	18	test	test	NOUN
dpc-1863	86	19	sets	set	NOUN
dpc-1863	86	20	(	(	PUNCT
dpc-1863	86	21	all	all	DET
dpc-1863	86	22	p	p	NOUN
dpc-1863	86	23	values	value	NOUN
dpc-1863	86	24	>	>	X
dpc-1863	86	25	0.17	0.17	NUM
dpc-1863	86	26	)	)	PUNCT
dpc-1863	86	27	.	.	PUNCT
dpc-1863	87	1	introducing	introduce	VERB
dpc-1863	87	2	artefacts	artefact	NOUN
dpc-1863	87	3	in	in	ADP
dpc-1863	87	4	the	the	DET
dpc-1863	87	5	training	training	NOUN
dpc-1863	87	6	data	datum	NOUN
dpc-1863	87	7	led	lead	VERB
dpc-1863	87	8	to	to	ADP
dpc-1863	87	9	biased	biased	ADJ
dpc-1863	87	10	results	result	NOUN
dpc-1863	87	11	in	in	ADP
dpc-1863	87	12	all	all	DET
dpc-1863	87	13	three	three	NUM
dpc-1863	87	14	examined	examine	VERB
dpc-1863	87	15	architectures	architecture	NOUN
dpc-1863	87	16	.	.	PUNCT
dpc-1863	88	1	artefacts	artefact	NOUN
dpc-1863	88	2	introduced	introduce	VERB
dpc-1863	88	3	figure	figure	NOUN
dpc-1863	88	4	3	3	NUM
dpc-1863	88	5	.	.	NOUN
dpc-1863	88	6	example	example	NOUN
dpc-1863	88	7	of	of	ADP
dpc-1863	88	8	automatically	automatically	ADV
dpc-1863	88	9	superimposed	superimpose	VERB
dpc-1863	88	10	artefacts	artefact	NOUN
dpc-1863	88	11	on	on	ADP
dpc-1863	88	12	a	a	DET
dpc-1863	88	13	dermatoscopic	dermatoscopic	ADJ
dpc-1863	88	14	image	image	NOUN
dpc-1863	88	15	.	.	PUNCT
dpc-1863	89	1	(	(	PUNCT
dpc-1863	89	2	a	a	X
dpc-1863	89	3	)	)	PUNCT
dpc-1863	89	4	in	in	ADP
dpc-1863	89	5	the	the	DET
dpc-1863	89	6	top	top	NOUN
dpc-1863	89	7	left	leave	VERB
dpc-1863	89	8	the	the	DET
dpc-1863	89	9	original	original	ADJ
dpc-1863	89	10	image	image	NOUN
dpc-1863	89	11	without	without	ADP
dpc-1863	89	12	artefact	artefact	NOUN
dpc-1863	89	13	is	be	AUX
dpc-1863	89	14	shown	show	VERB
dpc-1863	89	15	.	.	PUNCT
dpc-1863	90	1	the	the	DET
dpc-1863	90	2	other	other	ADJ
dpc-1863	90	3	3	3	NUM
dpc-1863	90	4	images	image	NOUN
dpc-1863	90	5	show	show	VERB
dpc-1863	90	6	the	the	DET
dpc-1863	90	7	lesion	lesion	NOUN
dpc-1863	90	8	with	with	ADP
dpc-1863	90	9	the	the	DET
dpc-1863	90	10	superimposed	superimpose	VERB
dpc-1863	90	11	artefacts	artefact	NOUN
dpc-1863	90	12	bubbles	bubble	NOUN
dpc-1863	90	13	(	(	PUNCT
dpc-1863	90	14	b	b	NOUN
dpc-1863	90	15	)	)	PUNCT
dpc-1863	90	16	,	,	PUNCT
dpc-1863	90	17	ink	ink	NOUN
dpc-1863	90	18	markings	marking	NOUN
dpc-1863	90	19	(	(	PUNCT
dpc-1863	90	20	c	c	NOUN
dpc-1863	90	21	)	)	PUNCT
dpc-1863	90	22	and	and	CCONJ
dpc-1863	90	23	a	a	DET
dpc-1863	90	24	ruler	ruler	NOUN
dpc-1863	90	25	(	(	PUNCT
dpc-1863	90	26	d	d	NOUN
dpc-1863	90	27	)	)	PUNCT
dpc-1863	90	28	.	.	PUNCT
dpc-1863	91	1	original	original	ADJ
dpc-1863	91	2	article	article	NOUN
dpc-1863	91	3	|	|	ADV
dpc-1863	91	4	dermatol	dermatol	ADJ
dpc-1863	91	5	pract	pract	NOUN
dpc-1863	91	6	concept	concept	NOUN
dpc-1863	91	7	.	.	PUNCT
dpc-1863	92	1	2022;12(3):e2022126	2022;12(3):e2022126	NUM
dpc-1863	92	2	5	5	NUM
dpc-1863	92	3	into	into	ADP
dpc-1863	92	4	all	all	DET
dpc-1863	92	5	images	image	NOUN
dpc-1863	92	6	of	of	ADP
dpc-1863	92	7	the	the	DET
dpc-1863	92	8	melanocytic	melanocytic	ADJ
dpc-1863	92	9	nevi	nevi	ADJ
dpc-1863	92	10	class	class	NOUN
dpc-1863	92	11	during	during	ADP
dpc-1863	92	12	training	training	NOUN
dpc-1863	92	13	decreased	decrease	VERB
dpc-1863	92	14	recall	recall	NOUN
dpc-1863	92	15	values	value	NOUN
dpc-1863	92	16	on	on	ADP
dpc-1863	92	17	average	average	ADJ
dpc-1863	92	18	by	by	ADP
dpc-1863	92	19	0.218	0.218	NUM
dpc-1863	92	20	for	for	ADP
dpc-1863	92	21	resnet-34	resnet-34	PROPN
dpc-1863	92	22	,	,	PUNCT
dpc-1863	92	23	0.129	0.129	NUM
dpc-1863	92	24	for	for	ADP
dpc-1863	92	25	faster	fast	ADJ
dpc-1863	92	26	r	r	NOUN
dpc-1863	92	27	-	-	PUNCT
dpc-1863	92	28	cnn	cnn	PROPN
dpc-1863	92	29	and	and	CCONJ
dpc-1863	92	30	by	by	ADP
dpc-1863	92	31	0.155	0.155	NUM
dpc-1863	92	32	for	for	ADP
dpc-1863	92	33	mask	mask	NOUN
dpc-1863	92	34	r	r	NOUN
dpc-1863	92	35	-	-	NOUN
dpc-1863	92	36	cnn	cnn	PROPN
dpc-1863	92	37	in	in	ADP
dpc-1863	92	38	comparison	comparison	NOUN
dpc-1863	92	39	to	to	ADP
dpc-1863	92	40	the	the	DET
dpc-1863	92	41	respective	respective	ADJ
dpc-1863	92	42	unbiased	unbiased	ADJ
dpc-1863	92	43	models	model	NOUN
dpc-1863	92	44	.	.	PUNCT
dpc-1863	93	1	reduction	reduction	NOUN
dpc-1863	93	2	in	in	ADP
dpc-1863	93	3	recall	recall	NOUN
dpc-1863	93	4	values	value	NOUN
dpc-1863	93	5	indicate	indicate	VERB
dpc-1863	93	6	that	that	SCONJ
dpc-1863	93	7	those	those	PRON
dpc-1863	93	8	are	be	AUX
dpc-1863	93	9	indeed	indeed	ADV
dpc-1863	93	10	biased	bias	VERB
dpc-1863	93	11	by	by	ADP
dpc-1863	93	12	artefacts	artefact	NOUN
dpc-1863	93	13	for	for	ADP
dpc-1863	93	14	specific	specific	ADJ
dpc-1863	93	15	classes	class	NOUN
dpc-1863	93	16	.	.	PUNCT
dpc-1863	94	1	this	this	DET
dpc-1863	94	2	effect	effect	NOUN
dpc-1863	94	3	was	be	AUX
dpc-1863	94	4	more	more	ADV
dpc-1863	94	5	apparent	apparent	ADJ
dpc-1863	94	6	the	the	PRON
dpc-1863	94	7	bigger	big	ADJ
dpc-1863	94	8	the	the	DET
dpc-1863	94	9	proportion	proportion	NOUN
dpc-1863	94	10	of	of	ADP
dpc-1863	94	11	biased	biased	ADJ
dpc-1863	94	12	samples	sample	NOUN
dpc-1863	94	13	in	in	ADP
dpc-1863	94	14	the	the	DET
dpc-1863	94	15	dataset	dataset	NOUN
dpc-1863	94	16	is	be	AUX
dpc-1863	94	17	.	.	PUNCT
dpc-1863	95	1	considering	consider	VERB
dpc-1863	95	2	the	the	DET
dpc-1863	95	3	fpr	fpr	NOUN
dpc-1863	95	4	and	and	CCONJ
dpc-1863	95	5	fnr	fnr	NOUN
dpc-1863	95	6	for	for	ADP
dpc-1863	95	7	specific	specific	ADJ
dpc-1863	95	8	classes	class	NOUN
dpc-1863	95	9	,	,	PUNCT
dpc-1863	95	10	a	a	DET
dpc-1863	95	11	selective	selective	ADJ
dpc-1863	95	12	bias	bias	NOUN
dpc-1863	95	13	towards	towards	ADP
dpc-1863	95	14	classes	class	NOUN
dpc-1863	95	15	that	that	PRON
dpc-1863	95	16	were	be	AUX
dpc-1863	95	17	corrupted	corrupt	VERB
dpc-1863	95	18	by	by	ADP
dpc-1863	95	19	artefacts	artefact	NOUN
dpc-1863	95	20	during	during	ADP
dpc-1863	95	21	training	training	NOUN
dpc-1863	95	22	could	could	AUX
dpc-1863	95	23	be	be	AUX
dpc-1863	95	24	observed	observe	VERB
dpc-1863	95	25	for	for	ADP
dpc-1863	95	26	all	all	DET
dpc-1863	95	27	three	three	NUM
dpc-1863	95	28	architectures	architecture	NOUN
dpc-1863	95	29	.	.	PUNCT
dpc-1863	96	1	the	the	DET
dpc-1863	96	2	increase	increase	NOUN
dpc-1863	96	3	in	in	ADP
dpc-1863	96	4	fpr	fpr	NOUN
dpc-1863	96	5	for	for	ADP
dpc-1863	96	6	the	the	DET
dpc-1863	96	7	class	class	NOUN
dpc-1863	96	8	with	with	ADP
dpc-1863	96	9	inserted	insert	VERB
dpc-1863	96	10	artefacts	artefact	NOUN
dpc-1863	96	11	during	during	ADP
dpc-1863	96	12	training	training	NOUN
dpc-1863	96	13	,	,	PUNCT
dpc-1863	96	14	and	and	CCONJ
dpc-1863	96	15	a	a	DET
dpc-1863	96	16	simultaneous	simultaneous	ADJ
dpc-1863	96	17	increase	increase	NOUN
dpc-1863	96	18	in	in	ADP
dpc-1863	96	19	fnr	fnr	PROPN
dpc-1863	96	20	for	for	ADP
dpc-1863	96	21	all	all	DET
dpc-1863	96	22	others	other	NOUN
dpc-1863	96	23	,	,	PUNCT
dpc-1863	96	24	in	in	ADP
dpc-1863	96	25	fact	fact	NOUN
dpc-1863	96	26	showed	show	VERB
dpc-1863	96	27	a	a	DET
dpc-1863	96	28	shift	shift	NOUN
dpc-1863	96	29	in	in	ADP
dpc-1863	96	30	classifications	classification	NOUN
dpc-1863	96	31	towards	towards	ADP
dpc-1863	96	32	the	the	DET
dpc-1863	96	33	biased	biased	ADJ
dpc-1863	96	34	class	class	NOUN
dpc-1863	96	35	.	.	PUNCT
dpc-1863	97	1	this	this	DET
dpc-1863	97	2	effect	effect	NOUN
dpc-1863	97	3	could	could	AUX
dpc-1863	97	4	not	not	PART
dpc-1863	97	5	be	be	AUX
dpc-1863	97	6	observed	observe	VERB
dpc-1863	97	7	if	if	SCONJ
dpc-1863	97	8	artefacts	artefact	NOUN
dpc-1863	97	9	were	be	AUX
dpc-1863	97	10	inserted	insert	VERB
dpc-1863	97	11	into	into	ADP
dpc-1863	97	12	none	none	NOUN
dpc-1863	97	13	or	or	CCONJ
dpc-1863	97	14	all	all	PRON
dpc-1863	97	15	of	of	ADP
dpc-1863	97	16	the	the	DET
dpc-1863	97	17	images	image	NOUN
dpc-1863	97	18	.	.	PUNCT
dpc-1863	98	1	when	when	SCONJ
dpc-1863	98	2	inspecting	inspect	VERB
dpc-1863	98	3	heat	heat	NOUN
dpc-1863	98	4	map	map	NOUN
dpc-1863	98	5	representations	representation	NOUN
dpc-1863	98	6	of	of	ADP
dpc-1863	98	7	the	the	DET
dpc-1863	98	8	gradcam	gradcam	NOUN
dpc-1863	98	9	we	we	PRON
dpc-1863	98	10	observed	observe	VERB
dpc-1863	98	11	that	that	SCONJ
dpc-1863	98	12	training	training	NOUN
dpc-1863	98	13	with	with	ADP
dpc-1863	98	14	artefacts	artefact	NOUN
dpc-1863	98	15	shifted	shift	VERB
dpc-1863	98	16	the	the	DET
dpc-1863	98	17	attention	attention	NOUN
dpc-1863	98	18	of	of	ADP
dpc-1863	98	19	the	the	DET
dpc-1863	98	20	object	object	NOUN
dpc-1863	98	21	detection	detection	NOUN
dpc-1863	98	22	and	and	CCONJ
dpc-1863	98	23	instance	instance	NOUN
dpc-1863	98	24	segmentation	segmentation	NOUN
dpc-1863	98	25	network	network	NOUN
dpc-1863	98	26	away	away	ADV
dpc-1863	98	27	from	from	ADP
dpc-1863	98	28	the	the	DET
dpc-1863	98	29	artefact	artefact	NOUN
dpc-1863	98	30	itself	itself	PRON
dpc-1863	98	31	towards	towards	ADP
dpc-1863	98	32	areas	area	NOUN
dpc-1863	98	33	of	of	ADP
dpc-1863	98	34	the	the	DET
dpc-1863	98	35	lesion	lesion	NOUN
dpc-1863	98	36	trained	train	VERB
dpc-1863	98	37	without	without	ADP
dpc-1863	98	38	inserted	insert	VERB
dpc-1863	98	39	artefacts	artefact	NOUN
dpc-1863	98	40	trained	train	VERB
dpc-1863	98	41	with	with	ADP
dpc-1863	98	42	inserted	insert	VERB
dpc-1863	98	43	artefacts	artefact	NOUN
dpc-1863	98	44	input	input	VERB
dpc-1863	98	45	a	a	DET
dpc-1863	98	46	b	b	NOUN
dpc-1863	98	47	c	c	NOUN
dpc-1863	98	48	d	d	NOUN
dpc-1863	98	49	hgfe	hgfe	NOUN
dpc-1863	98	50	lkji	lkji	NOUN
dpc-1863	98	51	resnet-34	resnet-34	PROPN
dpc-1863	98	52	faster	fast	ADV
dpc-1863	98	53	r	r	NOUN
dpc-1863	98	54	-	-	PUNCT
dpc-1863	98	55	cnn	cnn	PROPN
dpc-1863	98	56	mask	mask	NOUN
dpc-1863	98	57	r	r	PROPN
dpc-1863	98	58	-	-	PUNCT
dpc-1863	98	59	cnn	cnn	PROPN
dpc-1863	98	60	input	input	NOUN
dpc-1863	98	61	resnet-34	resnet-34	PROPN
dpc-1863	98	62	faster	fast	ADV
dpc-1863	98	63	r	r	NOUN
dpc-1863	98	64	-	-	PUNCT
dpc-1863	98	65	cnn	cnn	PROPN
dpc-1863	98	66	mask	mask	NOUN
dpc-1863	98	67	r	r	PROPN
dpc-1863	98	68	-	-	PUNCT
dpc-1863	98	69	cnn	cnn	PROPN
dpc-1863	98	70	o	o	NOUN
dpc-1863	98	71	pnm	pnm	PROPN
dpc-1863	98	72	figure	figure	NOUN
dpc-1863	98	73	5	5	NUM
dpc-1863	98	74	.	.	PUNCT
dpc-1863	98	75	grad	grad	NOUN
dpc-1863	98	76	-	-	PUNCT
dpc-1863	98	77	cam	cam	NOUN
dpc-1863	98	78	for	for	ADP
dpc-1863	98	79	used	use	VERB
dpc-1863	98	80	network	network	NOUN
dpc-1863	98	81	architectures	architecture	NOUN
dpc-1863	98	82	.	.	PUNCT
dpc-1863	99	1	the	the	DET
dpc-1863	99	2	first	first	ADJ
dpc-1863	99	3	column	column	NOUN
dpc-1863	99	4	shows	show	VERB
dpc-1863	99	5	the	the	DET
dpc-1863	99	6	input	input	NOUN
dpc-1863	99	7	image	image	NOUN
dpc-1863	99	8	for	for	ADP
dpc-1863	99	9	the	the	DET
dpc-1863	99	10	corresponding	corresponding	ADJ
dpc-1863	99	11	row	row	NOUN
dpc-1863	99	12	,	,	PUNCT
dpc-1863	99	13	in	in	ADP
dpc-1863	99	14	its	its	PRON
dpc-1863	99	15	original	original	ADJ
dpc-1863	99	16	form	form	NOUN
dpc-1863	99	17	(	(	PUNCT
dpc-1863	99	18	top	top	NOUN
dpc-1863	99	19	)	)	PUNCT
dpc-1863	99	20	and	and	CCONJ
dpc-1863	99	21	with	with	ADP
dpc-1863	99	22	bubble	bubble	NOUN
dpc-1863	99	23	artefacts	artefact	NOUN
dpc-1863	99	24	inserted	insert	VERB
dpc-1863	99	25	(	(	PUNCT
dpc-1863	99	26	bottom	bottom	NOUN
dpc-1863	99	27	)	)	PUNCT
dpc-1863	99	28	.	.	PUNCT
dpc-1863	100	1	grad	grad	NOUN
dpc-1863	100	2	-	-	PUNCT
dpc-1863	100	3	cam	cam	NOUN
dpc-1863	100	4	heatmaps	heatmap	NOUN
dpc-1863	100	5	show	show	VERB
dpc-1863	100	6	the	the	DET
dpc-1863	100	7	resnet-34	resnet-34	PROPN
dpc-1863	100	8	increases	increase	VERB
dpc-1863	100	9	attention	attention	NOUN
dpc-1863	100	10	towards	towards	ADP
dpc-1863	100	11	the	the	DET
dpc-1863	100	12	bubble	bubble	NOUN
dpc-1863	100	13	-	-	PUNCT
dpc-1863	100	14	area	area	NOUN
dpc-1863	100	15	after	after	ADP
dpc-1863	100	16	training	train	VERB
dpc-1863	100	17	with	with	ADP
dpc-1863	100	18	artefacts	artefact	NOUN
dpc-1863	100	19	(	(	PUNCT
dpc-1863	100	20	n	n	CCONJ
dpc-1863	100	21	)	)	PUNCT
dpc-1863	100	22	,	,	PUNCT
dpc-1863	100	23	where	where	SCONJ
dpc-1863	100	24	the	the	DET
dpc-1863	100	25	faster	fast	ADJ
dpc-1863	100	26	r	r	NOUN
dpc-1863	100	27	-	-	PUNCT
dpc-1863	100	28	cnn	cnn	PROPN
dpc-1863	100	29	network	network	NOUN
dpc-1863	100	30	loses	lose	VERB
dpc-1863	100	31	its	its	PRON
dpc-1863	100	32	initial	initial	ADJ
dpc-1863	100	33	attention	attention	NOUN
dpc-1863	100	34	towards	towards	ADP
dpc-1863	100	35	the	the	DET
dpc-1863	100	36	artefact	artefact	NOUN
dpc-1863	100	37	(	(	PUNCT
dpc-1863	100	38	g	g	NOUN
dpc-1863	100	39	)	)	PUNCT
dpc-1863	100	40	afterwards	afterwards	ADV
dpc-1863	100	41	(	(	PUNCT
dpc-1863	100	42	o	o	NOUN
dpc-1863	100	43	)	)	PUNCT
dpc-1863	100	44	.	.	PUNCT
dpc-1863	101	1	the	the	DET
dpc-1863	101	2	mask	mask	NOUN
dpc-1863	101	3	r	r	NOUN
dpc-1863	101	4	-	-	PUNCT
dpc-1863	101	5	cnn	cnn	PROPN
dpc-1863	101	6	architecture	architecture	NOUN
dpc-1863	101	7	seems	seem	VERB
dpc-1863	101	8	to	to	PART
dpc-1863	101	9	ignore	ignore	VERB
dpc-1863	101	10	the	the	DET
dpc-1863	101	11	artefact	artefact	NOUN
dpc-1863	101	12	throughout	throughout	ADP
dpc-1863	101	13	(	(	PUNCT
dpc-1863	101	14	h	h	NOUN
dpc-1863	101	15	and	and	CCONJ
dpc-1863	101	16	p	p	NOUN
dpc-1863	101	17	)	)	PUNCT
dpc-1863	101	18	.	.	PUNCT
dpc-1863	102	1	black	black	ADJ
dpc-1863	102	2	boxes	box	NOUN
dpc-1863	102	3	denote	denote	VERB
dpc-1863	102	4	positions	position	NOUN
dpc-1863	102	5	of	of	ADP
dpc-1863	102	6	inserted	insert	VERB
dpc-1863	102	7	bubble	bubble	NOUN
dpc-1863	102	8	artefacts	artefact	NOUN
dpc-1863	102	9	.	.	PUNCT
dpc-1863	103	1	0.00	0.00	NUM
dpc-1863	103	2	–	–	SYM
dpc-1863	103	3	0.02	0.02	NUM
dpc-1863	103	4	–	–	SYM
dpc-1863	103	5	0.04	0.04	NUM
dpc-1863	103	6	mask	mask	NOUN
dpc-1863	103	7	r	r	NOUN
dpc-1863	103	8	-	-	PUNCT
dpc-1863	103	9	cnn	cnn	NOUN
dpc-1863	103	10	resnet-34	resnet-34	PROPN
dpc-1863	103	11	faster	fast	ADV
dpc-1863	103	12	r	r	X
dpc-1863	103	13	-	-	PUNCT
dpc-1863	103	14	cnn	cnn	PROPN
dpc-1863	103	15	0.0	0.0	NUM
dpc-1863	103	16	0.2	0.2	NUM
dpc-1863	103	17	0.4	0.4	NUM
dpc-1863	103	18	0.6	0.6	NUM
dpc-1863	103	19	0.8	0.8	NUM
dpc-1863	103	20	1.0	1.0	NUM
dpc-1863	103	21	proportion	proportion	NOUN
dpc-1863	103	22	of	of	ADP
dpc-1863	103	23	artefacts	artefact	NOUN
dpc-1863	103	24	in	in	ADP
dpc-1863	103	25	the	the	DET
dpc-1863	103	26	test	test	NOUN
dpc-1863	103	27	set	set	VERB
dpc-1863	103	28	ch	ch	PROPN
dpc-1863	103	29	an	an	DET
dpc-1863	103	30	ge	ge	PROPN
dpc-1863	103	31	in	in	ADP
dpc-1863	103	32	p	p	PROPN
dpc-1863	103	33	r	r	NOUN
dpc-1863	103	34	a	a	DET
dpc-1863	103	35	u	u	NOUN
dpc-1863	103	36	c	c	PROPN
dpc-1863	103	37	–	–	PUNCT
dpc-1863	103	38	0.06	0.06	NUM
dpc-1863	103	39	figure	figure	NOUN
dpc-1863	103	40	4	4	NUM
dpc-1863	103	41	.	.	PUNCT
dpc-1863	103	42	neural	neural	ADJ
dpc-1863	103	43	networks	network	NOUN
dpc-1863	103	44	show	show	VERB
dpc-1863	103	45	different	different	ADJ
dpc-1863	103	46	robustness	robustness	NOUN
dpc-1863	103	47	to	to	ADP
dpc-1863	103	48	inserted	insert	VERB
dpc-1863	103	49	artefacts	artefact	NOUN
dpc-1863	103	50	on	on	ADP
dpc-1863	103	51	the	the	DET
dpc-1863	103	52	test	test	NOUN
dpc-1863	103	53	set	set	VERB
dpc-1863	103	54	.	.	PUNCT
dpc-1863	104	1	precision	precision	NOUN
dpc-1863	104	2	recall	recall	PROPN
dpc-1863	104	3	curve	curve	NOUN
dpc-1863	104	4	(	(	PUNCT
dpc-1863	104	5	pr	pr	NOUN
dpc-1863	104	6	-	-	PUNCT
dpc-1863	104	7	auc	auc	NOUN
dpc-1863	104	8	)	)	PUNCT
dpc-1863	104	9	as	as	SCONJ
dpc-1863	104	10	achieved	achieve	VERB
dpc-1863	104	11	by	by	ADP
dpc-1863	104	12	training	training	NOUN
dpc-1863	104	13	without	without	ADP
dpc-1863	104	14	additional	additional	ADJ
dpc-1863	104	15	artefacts	artefact	NOUN
dpc-1863	104	16	in	in	ADP
dpc-1863	104	17	the	the	DET
dpc-1863	104	18	train	train	NOUN
dpc-1863	104	19	and	and	CCONJ
dpc-1863	104	20	test	test	NOUN
dpc-1863	104	21	set	set	VERB
dpc-1863	104	22	was	be	AUX
dpc-1863	104	23	used	use	VERB
dpc-1863	104	24	as	as	ADP
dpc-1863	104	25	the	the	DET
dpc-1863	104	26	baseline	baseline	NOUN
dpc-1863	104	27	(	(	PUNCT
dpc-1863	104	28	0	0	NUM
dpc-1863	104	29	%	%	NOUN
dpc-1863	104	30	)	)	PUNCT
dpc-1863	104	31	.	.	PUNCT
dpc-1863	105	1	with	with	ADP
dpc-1863	105	2	increasing	increase	VERB
dpc-1863	105	3	proportion	proportion	NOUN
dpc-1863	105	4	of	of	ADP
dpc-1863	105	5	inserted	insert	VERB
dpc-1863	105	6	artefacts	artefact	NOUN
dpc-1863	105	7	,	,	PUNCT
dpc-1863	105	8	pr	pr	NOUN
dpc-1863	105	9	-	-	PUNCT
dpc-1863	105	10	auc	auc	NOUN
dpc-1863	105	11	decreases	decrease	NOUN
dpc-1863	105	12	for	for	ADP
dpc-1863	105	13	resnet-34	resnet-34	PROPN
dpc-1863	105	14	(	(	PUNCT
dpc-1863	105	15	blue	blue	ADJ
dpc-1863	105	16	)	)	PUNCT
dpc-1863	105	17	and	and	CCONJ
dpc-1863	105	18	faster	fast	ADJ
dpc-1863	105	19	r	r	X
dpc-1863	105	20	-	-	PUNCT
dpc-1863	105	21	cnn	cnn	PROPN
dpc-1863	105	22	(	(	PUNCT
dpc-1863	105	23	green	green	ADJ
dpc-1863	105	24	)	)	PUNCT
dpc-1863	105	25	,	,	PUNCT
dpc-1863	105	26	but	but	CCONJ
dpc-1863	105	27	almost	almost	ADV
dpc-1863	105	28	not	not	PART
dpc-1863	105	29	for	for	ADP
dpc-1863	105	30	mask	mask	NOUN
dpc-1863	105	31	r	r	NOUN
dpc-1863	105	32	-	-	PUNCT
dpc-1863	105	33	cnn	cnn	PROPN
dpc-1863	105	34	(	(	PUNCT
dpc-1863	105	35	purple	purple	ADJ
dpc-1863	105	36	)	)	PUNCT
dpc-1863	105	37	.	.	PUNCT
dpc-1863	106	1	shaded	shade	VERB
dpc-1863	106	2	areas	area	NOUN
dpc-1863	106	3	denote	denote	VERB
dpc-1863	106	4	95%-confidence	95%-confidence	NUM
dpc-1863	106	5	intervals	interval	NOUN
dpc-1863	106	6	.	.	PUNCT
dpc-1863	107	1	(	(	PUNCT
dpc-1863	107	2	figure	figure	NOUN
dpc-1863	107	3	5	5	NUM
dpc-1863	107	4	)	)	PUNCT
dpc-1863	107	5	.	.	PUNCT
dpc-1863	108	1	these	these	DET
dpc-1863	108	2	mappings	mapping	NOUN
dpc-1863	108	3	indicate	indicate	VERB
dpc-1863	108	4	an	an	DET
dpc-1863	108	5	increase	increase	NOUN
dpc-1863	108	6	in	in	ADP
dpc-1863	108	7	robustness	robustness	NOUN
dpc-1863	108	8	against	against	ADP
dpc-1863	108	9	these	these	DET
dpc-1863	108	10	very	very	ADJ
dpc-1863	108	11	artefacts	artefact	NOUN
dpc-1863	108	12	for	for	ADP
dpc-1863	108	13	faster	fast	ADJ
dpc-1863	108	14	and	and	CCONJ
dpc-1863	108	15	mask	mask	VERB
dpc-1863	108	16	r	r	NOUN
dpc-1863	108	17	-	-	PUNCT
dpc-1863	108	18	cnn	cnn	PROPN
dpc-1863	108	19	models	model	NOUN
dpc-1863	108	20	,	,	PUNCT
dpc-1863	108	21	if	if	SCONJ
dpc-1863	108	22	trained	train	VERB
dpc-1863	108	23	with	with	ADP
dpc-1863	108	24	inserted	insert	VERB
dpc-1863	108	25	artefacts	artefact	NOUN
dpc-1863	108	26	in	in	ADP
dpc-1863	108	27	the	the	DET
dpc-1863	108	28	dataset	dataset	NOUN
dpc-1863	108	29	.	.	PUNCT
dpc-1863	109	1	6	6	NUM
dpc-1863	109	2	original	original	ADJ
dpc-1863	109	3	article	article	NOUN
dpc-1863	109	4	|	|	ADV
dpc-1863	109	5	dermatol	dermatol	ADJ
dpc-1863	109	6	pract	pract	NOUN
dpc-1863	109	7	concept	concept	NOUN
dpc-1863	109	8	.	.	PUNCT
dpc-1863	110	1	2022;12(3):e2022126	2022;12(3):e2022126	NUM
dpc-1863	110	2	conclusions	conclusion	NOUN
dpc-1863	110	3	we	we	PRON
dpc-1863	110	4	compared	compare	VERB
dpc-1863	110	5	representatives	representative	NOUN
dpc-1863	110	6	of	of	ADP
dpc-1863	110	7	three	three	NUM
dpc-1863	110	8	neural	neural	ADJ
dpc-1863	110	9	network	network	NOUN
dpc-1863	110	10	architectures	architecture	NOUN
dpc-1863	110	11	to	to	PART
dpc-1863	110	12	classify	classify	VERB
dpc-1863	110	13	lesions	lesion	NOUN
dpc-1863	110	14	in	in	ADP
dpc-1863	110	15	dermatoscopic	dermatoscopic	NOUN
dpc-1863	110	16	images	image	NOUN
dpc-1863	110	17	in	in	ADP
dpc-1863	110	18	regard	regard	NOUN
dpc-1863	110	19	to	to	ADP
dpc-1863	110	20	their	their	PRON
dpc-1863	110	21	robustness	robustness	NOUN
dpc-1863	110	22	against	against	ADP
dpc-1863	110	23	artefacts	artefact	NOUN
dpc-1863	110	24	.	.	PUNCT
dpc-1863	111	1	although	although	SCONJ
dpc-1863	111	2	as	as	ADP
dpc-1863	111	3	a	a	DET
dpc-1863	111	4	limitation	limitation	NOUN
dpc-1863	111	5	the	the	DET
dpc-1863	111	6	baseline	baseline	ADJ
dpc-1863	111	7	performance	performance	NOUN
dpc-1863	111	8	of	of	ADP
dpc-1863	111	9	the	the	DET
dpc-1863	111	10	examined	examine	VERB
dpc-1863	111	11	models	model	NOUN
dpc-1863	111	12	were	be	AUX
dpc-1863	111	13	not	not	PART
dpc-1863	111	14	the	the	DET
dpc-1863	111	15	same	same	ADJ
dpc-1863	111	16	,	,	PUNCT
dpc-1863	111	17	we	we	PRON
dpc-1863	111	18	found	find	VERB
dpc-1863	111	19	differences	difference	NOUN
dpc-1863	111	20	in	in	ADP
dpc-1863	111	21	their	their	PRON
dpc-1863	111	22	vulnerability	vulnerability	NOUN
dpc-1863	111	23	to	to	ADP
dpc-1863	111	24	performance	performance	NOUN
dpc-1863	111	25	changes	change	NOUN
dpc-1863	111	26	under	under	ADP
dpc-1863	111	27	the	the	DET
dpc-1863	111	28	influence	influence	NOUN
dpc-1863	111	29	of	of	ADP
dpc-1863	111	30	artefacts	artefact	NOUN
dpc-1863	111	31	.	.	PUNCT
dpc-1863	112	1	mask	mask	NOUN
dpc-1863	112	2	r	r	NOUN
dpc-1863	112	3	-	-	PUNCT
dpc-1863	112	4	cnn	cnn	PROPN
dpc-1863	112	5	tends	tend	VERB
dpc-1863	112	6	to	to	PART
dpc-1863	112	7	be	be	AUX
dpc-1863	112	8	the	the	DET
dpc-1863	112	9	most	most	ADV
dpc-1863	112	10	robust	robust	ADJ
dpc-1863	112	11	.	.	PUNCT
dpc-1863	113	1	the	the	DET
dpc-1863	113	2	influence	influence	NOUN
dpc-1863	113	3	on	on	ADP
dpc-1863	113	4	classification	classification	NOUN
dpc-1863	113	5	results	result	NOUN
dpc-1863	113	6	by	by	ADP
dpc-1863	113	7	artefacts	artefact	NOUN
dpc-1863	113	8	in	in	ADP
dpc-1863	113	9	test	test	NOUN
dpc-1863	113	10	images	image	NOUN
dpc-1863	113	11	can	can	AUX
dpc-1863	113	12	be	be	AUX
dpc-1863	113	13	reduced	reduce	VERB
dpc-1863	113	14	by	by	ADP
dpc-1863	113	15	augmenting	augment	VERB
dpc-1863	113	16	training	training	NOUN
dpc-1863	113	17	data	datum	NOUN
dpc-1863	113	18	with	with	ADP
dpc-1863	113	19	artificially	artificially	ADV
dpc-1863	113	20	superimposed	superimpose	VERB
dpc-1863	113	21	artefacts	artefact	NOUN
dpc-1863	113	22	for	for	ADP
dpc-1863	113	23	all	all	DET
dpc-1863	113	24	three	three	NUM
dpc-1863	113	25	architectures	architecture	NOUN
dpc-1863	113	26	.	.	PUNCT
dpc-1863	114	1	this	this	PRON
dpc-1863	114	2	is	be	AUX
dpc-1863	114	3	in	in	ADP
dpc-1863	114	4	line	line	NOUN
dpc-1863	114	5	with	with	ADP
dpc-1863	114	6	findings	finding	NOUN
dpc-1863	114	7	by	by	ADP
dpc-1863	114	8	maron	maron	PROPN
dpc-1863	114	9	et	et	PROPN
dpc-1863	114	10	al	al	PROPN
dpc-1863	114	11	,	,	PUNCT
dpc-1863	114	12	who	who	PRON
dpc-1863	114	13	reduced	reduce	VERB
dpc-1863	114	14	but	but	CCONJ
dpc-1863	114	15	not	not	PART
dpc-1863	114	16	eliminated	eliminate	VERB
dpc-1863	114	17	brittleness	brittleness	NOUN
dpc-1863	114	18	of	of	ADP
dpc-1863	114	19	their	their	PRON
dpc-1863	114	20	system	system	NOUN
dpc-1863	114	21	through	through	ADP
dpc-1863	114	22	data	datum	NOUN
dpc-1863	114	23	augmentation	augmentation	NOUN
dpc-1863	114	24	[	[	X
dpc-1863	114	25	25	25	NUM
dpc-1863	114	26	]	]	PUNCT
dpc-1863	114	27	.	.	PUNCT
dpc-1863	115	1	we	we	PRON
dpc-1863	115	2	anticipate	anticipate	VERB
dpc-1863	115	3	that	that	SCONJ
dpc-1863	115	4	automated	automate	VERB
dpc-1863	115	5	superimposition	superimposition	NOUN
dpc-1863	115	6	of	of	ADP
dpc-1863	115	7	artefacts	artefact	NOUN
dpc-1863	115	8	as	as	SCONJ
dpc-1863	115	9	presented	present	VERB
dpc-1863	115	10	here	here	ADV
dpc-1863	115	11	as	as	ADP
dpc-1863	115	12	a	a	DET
dpc-1863	115	13	further	further	ADJ
dpc-1863	115	14	evolution	evolution	NOUN
dpc-1863	115	15	of	of	ADP
dpc-1863	115	16	data	datum	NOUN
dpc-1863	115	17	augmentation	augmentation	NOUN
dpc-1863	115	18	,	,	PUNCT
dpc-1863	115	19	that	that	SCONJ
dpc-1863	115	20	together	together	ADV
dpc-1863	115	21	with	with	ADP
dpc-1863	115	22	integrating	integrate	VERB
dpc-1863	115	23	more	more	ADV
dpc-1863	115	24	diverse	diverse	ADJ
dpc-1863	115	25	variants	variant	NOUN
dpc-1863	115	26	,	,	PUNCT
dpc-1863	115	27	will	will	AUX
dpc-1863	115	28	enhance	enhance	VERB
dpc-1863	115	29	robustness	robustness	NOUN
dpc-1863	115	30	of	of	ADP
dpc-1863	115	31	automated	automate	VERB
dpc-1863	115	32	classifiers	classifier	NOUN
dpc-1863	115	33	and	and	CCONJ
dpc-1863	115	34	decision	decision	NOUN
dpc-1863	115	35	support	support	NOUN
dpc-1863	115	36	systems	system	NOUN
dpc-1863	115	37	further	far	ADV
dpc-1863	115	38	[	[	X
dpc-1863	115	39	26,27	26,27	NUM
dpc-1863	115	40	]	]	PUNCT
dpc-1863	115	41	.	.	PUNCT
dpc-1863	116	1	the	the	DET
dpc-1863	116	2	initial	initial	ADJ
dpc-1863	116	3	data	datum	NOUN
dpc-1863	116	4	,	,	PUNCT
dpc-1863	116	5	in	in	ADP
dpc-1863	116	6	our	our	PRON
dpc-1863	116	7	view	view	NOUN
dpc-1863	116	8	,	,	PUNCT
dpc-1863	116	9	warrants	warrant	NOUN
dpc-1863	116	10	more	more	ADJ
dpc-1863	116	11	in	in	ADP
dpc-1863	116	12	-	-	PUNCT
dpc-1863	116	13	depth	depth	NOUN
dpc-1863	116	14	follow	follow	NOUN
dpc-1863	116	15	up	up	ADP
dpc-1863	116	16	research	research	NOUN
dpc-1863	116	17	on	on	ADP
dpc-1863	116	18	this	this	DET
dpc-1863	116	19	topic	topic	NOUN
dpc-1863	116	20	,	,	PUNCT
dpc-1863	116	21	to	to	PART
dpc-1863	116	22	understand	understand	VERB
dpc-1863	116	23	which	which	DET
dpc-1863	116	24	approaches	approach	NOUN
dpc-1863	116	25	are	be	AUX
dpc-1863	116	26	the	the	DET
dpc-1863	116	27	most	most	ADV
dpc-1863	116	28	effective	effective	ADJ
dpc-1863	116	29	and	and	CCONJ
dpc-1863	116	30	efficient	efficient	ADJ
dpc-1863	116	31	.	.	PUNCT
dpc-1863	117	1	however	however	ADV
dpc-1863	117	2	,	,	PUNCT
dpc-1863	117	3	this	this	DET
dpc-1863	117	4	work	work	NOUN
dpc-1863	117	5	failed	fail	VERB
dpc-1863	117	6	to	to	PART
dpc-1863	117	7	find	find	VERB
dpc-1863	117	8	evidence	evidence	NOUN
dpc-1863	117	9	for	for	ADP
dpc-1863	117	10	a	a	DET
dpc-1863	117	11	clinically	clinically	ADV
dpc-1863	117	12	relevant	relevant	ADJ
dpc-1863	117	13	robustness	robustness	NOUN
dpc-1863	117	14	against	against	ADP
dpc-1863	117	15	artefacts	artefact	NOUN
dpc-1863	117	16	of	of	ADP
dpc-1863	117	17	instance	instance	NOUN
dpc-1863	117	18	segmentation	segmentation	NOUN
dpc-1863	117	19	for	for	ADP
dpc-1863	117	20	several	several	ADJ
dpc-1863	117	21	reasons	reason	NOUN
dpc-1863	117	22	.	.	PUNCT
dpc-1863	118	1	on	on	ADP
dpc-1863	118	2	the	the	DET
dpc-1863	118	3	one	one	NUM
dpc-1863	118	4	hand	hand	NOUN
dpc-1863	118	5	we	we	PRON
dpc-1863	118	6	used	use	VERB
dpc-1863	118	7	a	a	DET
dpc-1863	118	8	shallow	shallow	ADJ
dpc-1863	118	9	backbone	backbone	NOUN
dpc-1863	118	10	network	network	NOUN
dpc-1863	118	11	architecture	architecture	NOUN
dpc-1863	118	12	for	for	ADP
dpc-1863	118	13	our	our	PRON
dpc-1863	118	14	experiments	experiment	NOUN
dpc-1863	118	15	,	,	PUNCT
dpc-1863	118	16	even	even	ADV
dpc-1863	118	17	though	though	SCONJ
dpc-1863	118	18	current	current	ADJ
dpc-1863	118	19	research	research	NOUN
dpc-1863	118	20	and	and	CCONJ
dpc-1863	118	21	commercial	commercial	ADJ
dpc-1863	118	22	products	product	NOUN
dpc-1863	118	23	commonly	commonly	ADV
dpc-1863	118	24	use	use	VERB
dpc-1863	118	25	deeper	deep	ADJ
dpc-1863	118	26	models	model	NOUN
dpc-1863	118	27	,	,	PUNCT
dpc-1863	118	28	and	and	CCONJ
dpc-1863	118	29	an	an	DET
dpc-1863	118	30	increase	increase	NOUN
dpc-1863	118	31	in	in	ADP
dpc-1863	118	32	robustness	robustness	NOUN
dpc-1863	118	33	against	against	ADP
dpc-1863	118	34	image	image	NOUN
dpc-1863	118	35	distortions	distortion	NOUN
dpc-1863	118	36	has	have	AUX
dpc-1863	118	37	been	be	AUX
dpc-1863	118	38	demonstrated	demonstrate	VERB
dpc-1863	118	39	by	by	ADP
dpc-1863	118	40	others	other	NOUN
dpc-1863	118	41	with	with	ADP
dpc-1863	118	42	increased	increase	VERB
dpc-1863	118	43	backbone	backbone	NOUN
dpc-1863	118	44	capacity	capacity	NOUN
dpc-1863	118	45	[	[	X
dpc-1863	118	46	28	28	NUM
dpc-1863	118	47	]	]	PUNCT
dpc-1863	118	48	.	.	PUNCT
dpc-1863	119	1	we	we	PRON
dpc-1863	119	2	also	also	ADV
dpc-1863	119	3	used	use	VERB
dpc-1863	119	4	a	a	DET
dpc-1863	119	5	new	new	ADJ
dpc-1863	119	6	template	template	NOUN
dpc-1863	119	7	-	-	PUNCT
dpc-1863	119	8	based	base	VERB
dpc-1863	119	9	approach	approach	NOUN
dpc-1863	119	10	to	to	PART
dpc-1863	119	11	superimpose	superimpose	VERB
dpc-1863	119	12	artefacts	artefact	NOUN
dpc-1863	119	13	on	on	ADP
dpc-1863	119	14	images	image	NOUN
dpc-1863	119	15	.	.	PUNCT
dpc-1863	120	1	this	this	DET
dpc-1863	120	2	approach	approach	NOUN
dpc-1863	120	3	leaves	leave	VERB
dpc-1863	120	4	room	room	NOUN
dpc-1863	120	5	for	for	ADP
dpc-1863	120	6	improvement	improvement	NOUN
dpc-1863	120	7	with	with	ADP
dpc-1863	120	8	regard	regard	NOUN
dpc-1863	120	9	to	to	ADP
dpc-1863	120	10	the	the	DET
dpc-1863	120	11	number	number	NOUN
dpc-1863	120	12	of	of	ADP
dpc-1863	120	13	images	image	NOUN
dpc-1863	120	14	the	the	DET
dpc-1863	120	15	artefacts	artefact	NOUN
dpc-1863	120	16	are	be	AUX
dpc-1863	120	17	extracted	extract	VERB
dpc-1863	120	18	from	from	ADP
dpc-1863	120	19	,	,	PUNCT
dpc-1863	120	20	and	and	CCONJ
dpc-1863	120	21	a	a	DET
dpc-1863	120	22	detailed	detailed	ADJ
dpc-1863	120	23	analysis	analysis	NOUN
dpc-1863	120	24	on	on	ADP
dpc-1863	120	25	how	how	SCONJ
dpc-1863	120	26	different	different	ADJ
dpc-1863	120	27	artefact	artefact	ADJ
dpc-1863	120	28	types	type	NOUN
dpc-1863	120	29	affect	affect	VERB
dpc-1863	120	30	the	the	DET
dpc-1863	120	31	classification	classification	NOUN
dpc-1863	120	32	performance	performance	NOUN
dpc-1863	120	33	.	.	PUNCT
dpc-1863	121	1	alternatively	alternatively	ADV
dpc-1863	121	2	,	,	PUNCT
dpc-1863	121	3	lesions	lesion	NOUN
dpc-1863	121	4	with	with	ADP
dpc-1863	121	5	existing	exist	VERB
dpc-1863	121	6	artefacts	artefact	NOUN
dpc-1863	121	7	could	could	AUX
dpc-1863	121	8	be	be	AUX
dpc-1863	121	9	used	use	VERB
dpc-1863	121	10	after	after	ADP
dpc-1863	121	11	manual	manual	NOUN
dpc-1863	121	12	or	or	CCONJ
dpc-1863	121	13	automated	automate	VERB
dpc-1863	121	14	annotations	annotation	NOUN
dpc-1863	121	15	.	.	PUNCT
dpc-1863	122	1	references	reference	NOUN
dpc-1863	122	2	1	1	NUM
dpc-1863	122	3	.	.	PUNCT
dpc-1863	123	1	apalla	apalla	PROPN
dpc-1863	123	2	z	z	PROPN
dpc-1863	123	3	,	,	PUNCT
dpc-1863	123	4	lallas	lallas	PROPN
dpc-1863	123	5	a	a	PRON
dpc-1863	123	6	,	,	PUNCT
dpc-1863	123	7	sotiriou	sotiriou	ADJ
dpc-1863	123	8	e	e	NOUN
dpc-1863	123	9	,	,	PUNCT
dpc-1863	123	10	lazaridou	lazaridou	PROPN
dpc-1863	123	11	e	e	NOUN
dpc-1863	123	12	,	,	PUNCT
dpc-1863	123	13	ioannides	ioannide	NOUN
dpc-1863	123	14	d.	d.	PROPN
dpc-1863	123	15	epidemiological	epidemiological	ADJ
dpc-1863	123	16	trends	trend	NOUN
dpc-1863	123	17	in	in	ADP
dpc-1863	123	18	skin	skin	NOUN
dpc-1863	123	19	cancer	cancer	NOUN
dpc-1863	123	20	.	.	PUNCT
dpc-1863	124	1	dermatol	dermatol	NOUN
dpc-1863	124	2	pract	pract	PROPN
dpc-1863	124	3	concept	concept	NOUN
dpc-1863	124	4	.	.	PUNCT
dpc-1863	125	1	2017;7(2):1	2017;7(2):1	NUM
dpc-1863	125	2	-	-	PUNCT
dpc-1863	125	3	6	6	NUM
dpc-1863	125	4	.	.	PUNCT
dpc-1863	125	5	doi	doi	NOUN
dpc-1863	125	6	:	:	PUNCT
dpc-1863	125	7	10.5826	10.5826	NUM
dpc-1863	125	8	/	/	SYM
dpc-1863	125	9	dpc.0702a01	dpc.0702a01	NOUN
dpc-1863	125	10	.	.	PUNCT
dpc-1863	126	1	pmid	pmid	NOUN
dpc-1863	126	2	:	:	PUNCT
dpc-1863	126	3	28515985	28515985	NUM
dpc-1863	126	4	.	.	PUNCT
dpc-1863	127	1	pmcid	pmcid	NOUN
dpc-1863	127	2	:	:	PUNCT
dpc-1863	127	3	pmc5424654	pmc5424654	NOUN
dpc-1863	127	4	.	.	PUNCT
dpc-1863	128	1	2	2	X
dpc-1863	128	2	.	.	X
dpc-1863	128	3	balch	balch	PROPN
dpc-1863	128	4	cm	cm	PROPN
dpc-1863	128	5	,	,	PUNCT
dpc-1863	128	6	soong	soong	PROPN
dpc-1863	128	7	s	s	PROPN
dpc-1863	128	8	-	-	PUNCT
dpc-1863	128	9	j	j	PROPN
dpc-1863	128	10	,	,	PUNCT
dpc-1863	128	11	atkins	atkins	PROPN
dpc-1863	128	12	mb	mb	PROPN
dpc-1863	128	13	,	,	PUNCT
dpc-1863	128	14	et	et	PROPN
dpc-1863	128	15	al	al	PROPN
dpc-1863	128	16	.	.	PUNCT
dpc-1863	129	1	an	an	DET
dpc-1863	129	2	evidence	evidence	NOUN
dpc-1863	129	3	-	-	PUNCT
dpc-1863	129	4	based	base	VERB
dpc-1863	129	5	staging	staging	NOUN
dpc-1863	129	6	system	system	NOUN
dpc-1863	129	7	for	for	ADP
dpc-1863	129	8	cutaneous	cutaneous	ADJ
dpc-1863	129	9	melanoma	melanoma	NOUN
dpc-1863	129	10	.	.	PUNCT
dpc-1863	130	1	ca	can	AUX
dpc-1863	130	2	cancer	cancer	PROPN
dpc-1863	130	3	j	j	PROPN
dpc-1863	130	4	clin	clin	PROPN
dpc-1863	130	5	.	.	PUNCT
dpc-1863	131	1	2004;54(3):131	2004;54(3):131	NOUN
dpc-1863	131	2	-	-	SYM
dpc-1863	131	3	149	149	NUM
dpc-1863	131	4	;	;	PUNCT
dpc-1863	131	5	quiz	quiz	NOUN
dpc-1863	131	6	182	182	NUM
dpc-1863	131	7	-	-	SYM
dpc-1863	131	8	184	184	NUM
dpc-1863	131	9	.	.	PUNCT
dpc-1863	132	1	doi	doi	NOUN
dpc-1863	132	2	:	:	PUNCT
dpc-1863	132	3	10.3322	10.3322	NUM
dpc-1863	132	4	/	/	SYM
dpc-1863	132	5	canjclin.54.3.131	canjclin.54.3.131	NUM
dpc-1863	132	6	.	.	PUNCT
dpc-1863	133	1	pmid	pmid	NOUN
dpc-1863	133	2	:	:	PUNCT
dpc-1863	133	3	15195788	15195788	NUM
dpc-1863	133	4	.	.	PUNCT
dpc-1863	134	1	3	3	X
dpc-1863	134	2	.	.	X
dpc-1863	134	3	kittler	kittler	PROPN
dpc-1863	134	4	h	h	PROPN
dpc-1863	134	5	,	,	PUNCT
dpc-1863	134	6	pehamberger	pehamberger	ADJ
dpc-1863	134	7	h	h	NOUN
dpc-1863	134	8	,	,	PUNCT
dpc-1863	134	9	wolff	wolff	PROPN
dpc-1863	134	10	k	k	PROPN
dpc-1863	134	11	,	,	PUNCT
dpc-1863	134	12	binder	binder	NOUN
dpc-1863	134	13	m.	m.	NOUN
dpc-1863	134	14	diagnostic	diagnostic	ADJ
dpc-1863	134	15	accuracy	accuracy	NOUN
dpc-1863	134	16	of	of	ADP
dpc-1863	134	17	dermoscopy	dermoscopy	NOUN
dpc-1863	134	18	.	.	PUNCT
dpc-1863	135	1	lancet	lancet	PROPN
dpc-1863	135	2	oncol	oncol	PROPN
dpc-1863	135	3	.	.	PUNCT
dpc-1863	136	1	2002;3(3):159	2002;3(3):159	NUM
dpc-1863	136	2	-	-	SYM
dpc-1863	136	3	165	165	NUM
dpc-1863	136	4	.	.	PUNCT
dpc-1863	137	1	doi	doi	NOUN
dpc-1863	137	2	:	:	PUNCT
dpc-1863	137	3	10.1016	10.1016	NUM
dpc-1863	137	4	/	/	SYM
dpc-1863	137	5	s1470	s1470	PROPN
dpc-1863	137	6	-	-	PUNCT
dpc-1863	137	7	2045(02)00679	2045(02)00679	NUM
dpc-1863	137	8	-	-	PUNCT
dpc-1863	137	9	4	4	NUM
dpc-1863	137	10	.	.	PUNCT
dpc-1863	138	1	pmid	pmid	NOUN
dpc-1863	138	2	:	:	PUNCT
dpc-1863	138	3	11902502	11902502	NUM
dpc-1863	138	4	.	.	PUNCT
dpc-1863	139	1	4	4	X
dpc-1863	139	2	.	.	X
dpc-1863	139	3	korotkov	korotkov	PROPN
dpc-1863	139	4	k	k	PROPN
dpc-1863	139	5	,	,	PUNCT
dpc-1863	139	6	garcia	garcia	PROPN
dpc-1863	139	7	r.	r.	PROPN
dpc-1863	139	8	computerized	computerize	VERB
dpc-1863	139	9	analysis	analysis	NOUN
dpc-1863	139	10	of	of	ADP
dpc-1863	139	11	pigmented	pigmented	ADJ
dpc-1863	139	12	skin	skin	NOUN
dpc-1863	139	13	lesions	lesion	NOUN
dpc-1863	139	14	:	:	PUNCT
dpc-1863	139	15	a	a	DET
dpc-1863	139	16	review	review	NOUN
dpc-1863	139	17	.	.	PUNCT
dpc-1863	140	1	artif	artif	PROPN
dpc-1863	140	2	intell	intell	PROPN
dpc-1863	140	3	med	med	PROPN
dpc-1863	140	4	.	.	PUNCT
dpc-1863	141	1	2012;56(2):69	2012;56(2):69	NUM
dpc-1863	141	2	-	-	SYM
dpc-1863	141	3	90	90	NUM
dpc-1863	141	4	.	.	PUNCT
dpc-1863	142	1	doi	doi	NOUN
dpc-1863	142	2	:	:	PUNCT
dpc-1863	142	3	10.1016	10.1016	NUM
dpc-1863	142	4	/	/	SYM
dpc-1863	142	5	j.artmed.2012.08.002	j.artmed.2012.08.002	PROPN
dpc-1863	142	6	.	.	PUNCT
dpc-1863	143	1	pmid	pmid	NOUN
dpc-1863	143	2	:	:	PUNCT
dpc-1863	143	3	23063256	23063256	NUM
dpc-1863	143	4	.	.	PUNCT
dpc-1863	144	1	5	5	X
dpc-1863	144	2	.	.	X
dpc-1863	144	3	rubegni	rubegni	VERB
dpc-1863	144	4	p	p	NOUN
dpc-1863	144	5	,	,	PUNCT
dpc-1863	144	6	burroni	burroni	PROPN
dpc-1863	144	7	m	m	PROPN
dpc-1863	144	8	,	,	PUNCT
dpc-1863	144	9	cevenini	cevenini	VERB
dpc-1863	144	10	g	g	PROPN
dpc-1863	144	11	,	,	PUNCT
dpc-1863	144	12	et	et	PROPN
dpc-1863	144	13	al	al	PROPN
dpc-1863	144	14	.	.	PUNCT
dpc-1863	145	1	digital	digital	PROPN
dpc-1863	145	2	dermoscopy	dermoscopy	VERB
dpc-1863	145	3	analysis	analysis	NOUN
dpc-1863	145	4	and	and	CCONJ
dpc-1863	145	5	artificial	artificial	ADJ
dpc-1863	145	6	neural	neural	ADJ
dpc-1863	145	7	network	network	NOUN
dpc-1863	145	8	for	for	ADP
dpc-1863	145	9	the	the	DET
dpc-1863	145	10	differentiation	differentiation	NOUN
dpc-1863	145	11	of	of	ADP
dpc-1863	145	12	clinically	clinically	ADV
dpc-1863	145	13	atypical	atypical	ADJ
dpc-1863	145	14	pigmented	pigmented	ADJ
dpc-1863	145	15	skin	skin	NOUN
dpc-1863	145	16	lesions	lesion	NOUN
dpc-1863	145	17	:	:	PUNCT
dpc-1863	145	18	a	a	DET
dpc-1863	145	19	retrospective	retrospective	ADJ
dpc-1863	145	20	study	study	NOUN
dpc-1863	145	21	.	.	PUNCT
dpc-1863	146	1	j	j	PROPN
dpc-1863	146	2	invest	invest	VERB
dpc-1863	146	3	dermatol	dermatol	NOUN
dpc-1863	146	4	.	.	PUNCT
dpc-1863	147	1	2002;119(2):471	2002;119(2):471	PROPN
dpc-1863	147	2	-	-	SYM
dpc-1863	147	3	474	474	NUM
dpc-1863	147	4	.	.	PUNCT
dpc-1863	148	1	doi	doi	NOUN
dpc-1863	148	2	:	:	PUNCT
dpc-1863	148	3	10.1046	10.1046	NUM
dpc-1863	148	4	/	/	SYM
dpc-1863	148	5	j.15231747.2002.01835.x	j.15231747.2002.01835.x	PROPN
dpc-1863	148	6	.	.	PUNCT
dpc-1863	149	1	pmid	pmid	NOUN
dpc-1863	149	2	:	:	PUNCT
dpc-1863	149	3	12190872	12190872	NUM
dpc-1863	149	4	.	.	PUNCT
dpc-1863	150	1	6	6	NUM
dpc-1863	150	2	.	.	X
dpc-1863	150	3	esteva	esteva	PROPN
dpc-1863	150	4	a	a	PRON
dpc-1863	150	5	,	,	PUNCT
dpc-1863	150	6	kuprel	kuprel	NOUN
dpc-1863	150	7	b	b	NOUN
dpc-1863	150	8	,	,	PUNCT
dpc-1863	150	9	novoa	novoa	ADJ
dpc-1863	150	10	ra	ra	PROPN
dpc-1863	150	11	,	,	PUNCT
dpc-1863	150	12	et	et	PROPN
dpc-1863	150	13	al	al	PROPN
dpc-1863	150	14	.	.	PROPN
dpc-1863	150	15	dermatologist	dermatologist	NOUN
dpc-1863	150	16	-	-	PUNCT
dpc-1863	150	17	level	level	NOUN
dpc-1863	150	18	classification	classification	NOUN
dpc-1863	150	19	of	of	ADP
dpc-1863	150	20	skin	skin	NOUN
dpc-1863	150	21	cancer	cancer	NOUN
dpc-1863	150	22	with	with	ADP
dpc-1863	150	23	deep	deep	ADJ
dpc-1863	150	24	neural	neural	ADJ
dpc-1863	150	25	networks	network	NOUN
dpc-1863	150	26	.	.	PUNCT
dpc-1863	151	1	nature	nature	NOUN
dpc-1863	151	2	.	.	PUNCT
dpc-1863	152	1	2017;542(7639):115	2017;542(7639):115	NUM
dpc-1863	152	2	-	-	SYM
dpc-1863	152	3	118	118	NUM
dpc-1863	152	4	.	.	PUNCT
dpc-1863	153	1	doi	doi	NOUN
dpc-1863	153	2	:	:	PUNCT
dpc-1863	153	3	10.1038	10.1038	NUM
dpc-1863	153	4	/	/	SYM
dpc-1863	153	5	nature21056	nature21056	PROPN
dpc-1863	153	6	.	.	PUNCT
dpc-1863	154	1	pmid	pmid	NOUN
dpc-1863	154	2	:	:	PUNCT
dpc-1863	154	3	28117445	28117445	NUM
dpc-1863	154	4	.	.	PUNCT
dpc-1863	155	1	pmcid	pmcid	NOUN
dpc-1863	155	2	:	:	PUNCT
dpc-1863	156	1	pmc8382232	pmc8382232	NOUN
dpc-1863	156	2	.	.	PUNCT
dpc-1863	157	1	7	7	X
dpc-1863	157	2	.	.	X
dpc-1863	157	3	muñoz	muñoz	PROPN
dpc-1863	157	4	-	-	PUNCT
dpc-1863	157	5	lópez	lópez	PROPN
dpc-1863	157	6	c	c	NOUN
dpc-1863	157	7	,	,	PUNCT
dpc-1863	157	8	ramírez	ramírez	NOUN
dpc-1863	157	9	-	-	PUNCT
dpc-1863	157	10	cornejo	cornejo	PROPN
dpc-1863	157	11	c	c	PROPN
dpc-1863	157	12	,	,	PUNCT
dpc-1863	157	13	marchetti	marchetti	PROPN
dpc-1863	157	14	ma	ma	PROPN
dpc-1863	157	15	,	,	PUNCT
dpc-1863	157	16	et	et	PROPN
dpc-1863	157	17	al	al	PROPN
dpc-1863	157	18	.	.	PUNCT
dpc-1863	157	19	performance	performance	NOUN
dpc-1863	157	20	of	of	ADP
dpc-1863	157	21	a	a	DET
dpc-1863	157	22	deep	deep	ADJ
dpc-1863	157	23	neural	neural	ADJ
dpc-1863	157	24	network	network	NOUN
dpc-1863	157	25	in	in	ADP
dpc-1863	157	26	teledermatology	teledermatology	NOUN
dpc-1863	157	27	:	:	PUNCT
dpc-1863	157	28	a	a	DET
dpc-1863	157	29	single	single	ADJ
dpc-1863	157	30	-	-	PUNCT
dpc-1863	157	31	centre	centre	NOUN
dpc-1863	157	32	prospective	prospective	ADJ
dpc-1863	157	33	diagnostic	diagnostic	ADJ
dpc-1863	157	34	study	study	NOUN
dpc-1863	157	35	.	.	PUNCT
dpc-1863	158	1	j	j	PROPN
dpc-1863	158	2	eur	eur	PROPN
dpc-1863	158	3	acad	acad	PROPN
dpc-1863	158	4	dermatol	dermatol	NOUN
dpc-1863	158	5	venereol	venereol	NOUN
dpc-1863	158	6	.	.	PUNCT
dpc-1863	159	1	2021;35(2):546	2021;35(2):546	NUM
dpc-1863	159	2	-	-	PUNCT
dpc-1863	159	3	553	553	NUM
dpc-1863	159	4	.	.	PUNCT
dpc-1863	160	1	doi	doi	NOUN
dpc-1863	160	2	:	:	PUNCT
dpc-1863	160	3	10.1111	10.1111	NUM
dpc-1863	160	4	/	/	SYM
dpc-1863	160	5	jdv.16979	jdv.16979	NOUN
dpc-1863	160	6	.	.	PUNCT
dpc-1863	161	1	pmid	pmid	NOUN
dpc-1863	161	2	:	:	PUNCT
dpc-1863	161	3	33037709	33037709	NUM
dpc-1863	161	4	.	.	PUNCT
dpc-1863	162	1	pmcid	pmcid	NOUN
dpc-1863	162	2	:	:	PUNCT
dpc-1863	162	3	pmc8274350	pmc8274350	VERB
dpc-1863	162	4	.	.	PUNCT
dpc-1863	163	1	8	8	X
dpc-1863	163	2	.	.	X
dpc-1863	163	3	tschandl	tschandl	PROPN
dpc-1863	163	4	p	p	NOUN
dpc-1863	163	5	,	,	PUNCT
dpc-1863	163	6	rinner	rinner	NOUN
dpc-1863	163	7	c	c	NOUN
dpc-1863	163	8	,	,	PUNCT
dpc-1863	163	9	apalla	apalla	PROPN
dpc-1863	163	10	z	z	PROPN
dpc-1863	163	11	,	,	PUNCT
dpc-1863	163	12	et	et	PROPN
dpc-1863	163	13	al	al	PROPN
dpc-1863	163	14	.	.	PUNCT
dpc-1863	164	1	human	human	ADJ
dpc-1863	164	2	–	–	PUNCT
dpc-1863	164	3	computer	computer	NOUN
dpc-1863	164	4	collaboration	collaboration	NOUN
dpc-1863	164	5	for	for	ADP
dpc-1863	164	6	skin	skin	NOUN
dpc-1863	164	7	cancer	cancer	NOUN
dpc-1863	164	8	recognition	recognition	NOUN
dpc-1863	164	9	.	.	PUNCT
dpc-1863	165	1	nat	nat	PROPN
dpc-1863	165	2	med	me	VERB
dpc-1863	165	3	.	.	PUNCT
dpc-1863	166	1	2020;26(8):12291234	2020;26(8):12291234	X
dpc-1863	166	2	.	.	PUNCT
dpc-1863	166	3	doi	doi	NOUN
dpc-1863	166	4	:	:	PUNCT
dpc-1863	166	5	10.1038	10.1038	NUM
dpc-1863	166	6	/	/	SYM
dpc-1863	166	7	s41591	s41591	VERB
dpc-1863	166	8	-	-	PUNCT
dpc-1863	166	9	020	020	NUM
dpc-1863	166	10	-	-	PUNCT
dpc-1863	166	11	0942	0942	NUM
dpc-1863	166	12	-	-	PUNCT
dpc-1863	166	13	0	0	NUM
dpc-1863	166	14	.	.	PUNCT
dpc-1863	167	1	pmid	pmid	NOUN
dpc-1863	167	2	:	:	PUNCT
dpc-1863	167	3	32572267	32572267	NUM
dpc-1863	167	4	.	.	PUNCT
dpc-1863	168	1	9	9	X
dpc-1863	168	2	.	.	X
dpc-1863	168	3	fink	fink	PROPN
dpc-1863	168	4	c	c	PROPN
dpc-1863	168	5	,	,	PUNCT
dpc-1863	168	6	blum	blum	PROPN
dpc-1863	168	7	a	a	PROPN
dpc-1863	168	8	,	,	PUNCT
dpc-1863	168	9	buhl	buhl	PROPN
dpc-1863	168	10	t	t	PROPN
dpc-1863	168	11	,	,	PUNCT
dpc-1863	168	12	et	et	PROPN
dpc-1863	168	13	al	al	PROPN
dpc-1863	168	14	.	.	PUNCT
dpc-1863	169	1	diagnostic	diagnostic	ADJ
dpc-1863	169	2	performance	performance	NOUN
dpc-1863	169	3	of	of	ADP
dpc-1863	169	4	a	a	DET
dpc-1863	169	5	deep	deep	ADJ
dpc-1863	169	6	learning	learning	NOUN
dpc-1863	169	7	convolutional	convolutional	ADJ
dpc-1863	169	8	neural	neural	ADJ
dpc-1863	169	9	network	network	NOUN
dpc-1863	169	10	in	in	ADP
dpc-1863	169	11	the	the	DET
dpc-1863	169	12	differentiation	differentiation	NOUN
dpc-1863	169	13	of	of	ADP
dpc-1863	169	14	combined	combined	ADJ
dpc-1863	169	15	naevi	naevi	NOUN
dpc-1863	169	16	and	and	CCONJ
dpc-1863	169	17	melanomas	melanoma	NOUN
dpc-1863	169	18	.	.	PUNCT
dpc-1863	170	1	j	j	PROPN
dpc-1863	170	2	eur	eur	PROPN
dpc-1863	170	3	acad	acad	PROPN
dpc-1863	170	4	dermatol	dermatol	NOUN
dpc-1863	170	5	venereol	venereol	NOUN
dpc-1863	170	6	.	.	PUNCT
dpc-1863	171	1	2020;34(6):1355	2020;34(6):1355	NUM
dpc-1863	171	2	-	-	SYM
dpc-1863	171	3	1361	1361	NUM
dpc-1863	171	4	.	.	PUNCT
dpc-1863	172	1	doi	doi	NOUN
dpc-1863	172	2	:	:	PUNCT
dpc-1863	172	3	10.1111	10.1111	NUM
dpc-1863	172	4	/	/	SYM
dpc-1863	172	5	jdv.16165	jdv.16165	NOUN
dpc-1863	172	6	.	.	PUNCT
dpc-1863	173	1	pmid	pmid	NOUN
dpc-1863	173	2	:	:	PUNCT
dpc-1863	173	3	31856342	31856342	NUM
dpc-1863	173	4	.	.	PUNCT
dpc-1863	174	1	10	10	NUM
dpc-1863	174	2	.	.	PUNCT
dpc-1863	175	1	okuboyejo	okuboyejo	PROPN
dpc-1863	175	2	da	da	PROPN
dpc-1863	175	3	,	,	PUNCT
dpc-1863	175	4	olugbara	olugbara	NOUN
dpc-1863	175	5	oo	oo	INTJ
dpc-1863	175	6	.	.	PUNCT
dpc-1863	176	1	a	a	DET
dpc-1863	176	2	review	review	NOUN
dpc-1863	176	3	of	of	ADP
dpc-1863	176	4	prevalent	prevalent	ADJ
dpc-1863	176	5	methods	method	NOUN
dpc-1863	176	6	for	for	ADP
dpc-1863	176	7	automatic	automatic	ADJ
dpc-1863	176	8	skin	skin	NOUN
dpc-1863	176	9	lesion	lesion	NOUN
dpc-1863	176	10	diagnosis	diagnosis	NOUN
dpc-1863	176	11	.	.	PUNCT
dpc-1863	177	1	open	open	ADJ
dpc-1863	177	2	dermatol	dermatol	NOUN
dpc-1863	177	3	j.	j.	PROPN
dpc-1863	177	4	2018;12(1):14	2018;12(1):14	PROPN
dpc-1863	177	5	-	-	PROPN
dpc-1863	177	6	53	53	NUM
dpc-1863	177	7	.	.	PUNCT
dpc-1863	178	1	doi	doi	NOUN
dpc-1863	178	2	:	:	PUNCT
dpc-1863	178	3	10.2174/187437220181201014	10.2174/187437220181201014	NUM
dpc-1863	178	4	11	11	NUM
dpc-1863	178	5	.	.	PUNCT
dpc-1863	179	1	winkler	winkler	PROPN
dpc-1863	179	2	jk	jk	PROPN
dpc-1863	179	3	,	,	PUNCT
dpc-1863	179	4	fink	fink	PROPN
dpc-1863	179	5	c	c	PROPN
dpc-1863	179	6	,	,	PUNCT
dpc-1863	179	7	toberer	toberer	NOUN
dpc-1863	179	8	f	f	PROPN
dpc-1863	179	9	,	,	PUNCT
dpc-1863	179	10	et	et	PROPN
dpc-1863	179	11	al	al	PROPN
dpc-1863	179	12	.	.	PROPN
dpc-1863	179	13	association	association	PROPN
dpc-1863	179	14	between	between	ADP
dpc-1863	179	15	surgical	surgical	ADJ
dpc-1863	179	16	skin	skin	NOUN
dpc-1863	179	17	markings	marking	NOUN
dpc-1863	179	18	in	in	ADP
dpc-1863	179	19	dermoscopic	dermoscopic	ADJ
dpc-1863	179	20	images	image	NOUN
dpc-1863	179	21	and	and	CCONJ
dpc-1863	179	22	diagnostic	diagnostic	ADJ
dpc-1863	179	23	performance	performance	NOUN
dpc-1863	179	24	of	of	ADP
dpc-1863	179	25	a	a	DET
dpc-1863	179	26	deep	deep	ADJ
dpc-1863	179	27	learning	learning	NOUN
dpc-1863	179	28	convolutional	convolutional	ADJ
dpc-1863	179	29	neural	neural	ADJ
dpc-1863	179	30	network	network	NOUN
dpc-1863	179	31	for	for	ADP
dpc-1863	179	32	melanoma	melanoma	NOUN
dpc-1863	179	33	recognition	recognition	NOUN
dpc-1863	179	34	.	.	PUNCT
dpc-1863	180	1	jama	jama	PROPN
dpc-1863	180	2	dermatol	dermatol	NOUN
dpc-1863	180	3	.	.	PUNCT
dpc-1863	181	1	2019;155(10):11351141	2019;155(10):11351141	X
dpc-1863	181	2	.	.	PUNCT
dpc-1863	182	1	doi	doi	NOUN
dpc-1863	182	2	:	:	PUNCT
dpc-1863	182	3	10.1001	10.1001	NUM
dpc-1863	182	4	/	/	SYM
dpc-1863	182	5	jamadermatol.2019.1735	jamadermatol.2019.1735	NOUN
dpc-1863	182	6	.	.	PUNCT
dpc-1863	183	1	pmid	pmid	NOUN
dpc-1863	183	2	:	:	PUNCT
dpc-1863	183	3	31411641	31411641	NUM
dpc-1863	183	4	.	.	PUNCT
dpc-1863	184	1	pmcid	pmcid	NOUN
dpc-1863	184	2	:	:	PUNCT
dpc-1863	184	3	pmc6694463	pmc6694463	ADJ
dpc-1863	184	4	.	.	PUNCT
dpc-1863	185	1	12	12	NUM
dpc-1863	185	2	.	.	PUNCT
dpc-1863	186	1	navarrete	navarrete	ADJ
dpc-1863	186	2	-	-	PUNCT
dpc-1863	186	3	dechent	dechent	NOUN
dpc-1863	186	4	c	c	NOUN
dpc-1863	186	5	,	,	PUNCT
dpc-1863	186	6	dusza	dusza	PROPN
dpc-1863	186	7	sw	sw	PROPN
dpc-1863	186	8	,	,	PUNCT
dpc-1863	186	9	liopyris	liopyris	PROPN
dpc-1863	186	10	k	k	PROPN
dpc-1863	186	11	,	,	PUNCT
dpc-1863	186	12	marghoob	marghoob	VERB
dpc-1863	186	13	aa	aa	NOUN
dpc-1863	186	14	,	,	PUNCT
dpc-1863	186	15	halpern	halpern	PROPN
dpc-1863	186	16	ac	ac	PROPN
dpc-1863	186	17	,	,	PUNCT
dpc-1863	186	18	marchetti	marchetti	PROPN
dpc-1863	186	19	ma	ma	PROPN
dpc-1863	186	20	.	.	PROPN
dpc-1863	186	21	automated	automate	VERB
dpc-1863	186	22	dermatological	dermatological	ADJ
dpc-1863	186	23	diagnosis	diagnosis	NOUN
dpc-1863	186	24	:	:	PUNCT
dpc-1863	186	25	hype	hype	NOUN
dpc-1863	186	26	or	or	CCONJ
dpc-1863	186	27	reality	reality	NOUN
dpc-1863	186	28	?	?	PUNCT
dpc-1863	187	1	j	j	PROPN
dpc-1863	187	2	invest	invest	VERB
dpc-1863	187	3	dermatol	dermatol	NOUN
dpc-1863	187	4	.	.	PUNCT
dpc-1863	188	1	2018;138(10):22772279	2018;138(10):22772279	NUM
dpc-1863	188	2	.	.	PUNCT
dpc-1863	188	3	doi	doi	PROPN
dpc-1863	188	4	:	:	PUNCT
dpc-1863	188	5	10.1016	10.1016	NUM
dpc-1863	188	6	/	/	SYM
dpc-1863	188	7	j.jid.2018.04.040	j.jid.2018.04.040	NOUN
dpc-1863	188	8	.	.	PUNCT
dpc-1863	189	1	pmid	pmid	NOUN
dpc-1863	189	2	:	:	PUNCT
dpc-1863	189	3	29864435	29864435	NUM
dpc-1863	189	4	.	.	PUNCT
dpc-1863	190	1	pmcid	pmcid	ADJ
dpc-1863	190	2	:	:	PUNCT
dpc-1863	190	3	pmc7701995	pmc7701995	ADJ
dpc-1863	190	4	.	.	PUNCT
dpc-1863	191	1	13	13	NUM
dpc-1863	191	2	.	.	PUNCT
dpc-1863	192	1	ren	ren	PROPN
dpc-1863	192	2	s	s	PROPN
dpc-1863	192	3	,	,	PUNCT
dpc-1863	192	4	he	he	PRON
dpc-1863	192	5	k	k	NOUN
dpc-1863	192	6	,	,	PUNCT
dpc-1863	192	7	girshick	girshick	ADJ
dpc-1863	192	8	r	r	NOUN
dpc-1863	192	9	,	,	PUNCT
dpc-1863	192	10	sun	sun	PROPN
dpc-1863	192	11	j.	j.	PROPN
dpc-1863	192	12	faster	fast	ADV
dpc-1863	192	13	r	r	PROPN
dpc-1863	192	14	-	-	PUNCT
dpc-1863	192	15	cnn	cnn	NOUN
dpc-1863	192	16	:	:	PUNCT
dpc-1863	192	17	towards	towards	ADP
dpc-1863	192	18	real	real	ADJ
dpc-1863	192	19	-	-	PUNCT
dpc-1863	192	20	time	time	NOUN
dpc-1863	192	21	object	object	NOUN
dpc-1863	192	22	detection	detection	NOUN
dpc-1863	192	23	with	with	ADP
dpc-1863	192	24	region	region	NOUN
dpc-1863	192	25	proposal	proposal	NOUN
dpc-1863	192	26	networks	network	NOUN
dpc-1863	192	27	.	.	PUNCT
dpc-1863	193	1	ieee	ieee	NOUN
dpc-1863	193	2	transactions	transaction	NOUN
dpc-1863	193	3	on	on	ADP
dpc-1863	193	4	pattern	pattern	NOUN
dpc-1863	193	5	analysis	analysis	NOUN
dpc-1863	193	6	and	and	CCONJ
dpc-1863	193	7	machine	machine	NOUN
dpc-1863	193	8	intelligence	intelligence	NOUN
dpc-1863	193	9	.	.	PUNCT
dpc-1863	194	1	2017;39(6):1137	2017;39(6):1137	NUM
dpc-1863	194	2	-	-	SYM
dpc-1863	194	3	1149	1149	NUM
dpc-1863	194	4	.	.	PUNCT
dpc-1863	194	5	.	.	PUNCT
dpc-1863	195	1	doi	doi	NOUN
dpc-1863	195	2	:	:	PUNCT
dpc-1863	195	3	10.1109	10.1109	NUM
dpc-1863	195	4	/	/	SYM
dpc-1863	195	5	tpami.2016.2577031	tpami.2016.2577031	NOUN
dpc-1863	195	6	.	.	PUNCT
dpc-1863	196	1	pmid	pmid	NOUN
dpc-1863	196	2	:	:	PUNCT
dpc-1863	196	3	27295650	27295650	NUM
dpc-1863	196	4	.	.	PUNCT
dpc-1863	197	1	14	14	NUM
dpc-1863	197	2	.	.	PUNCT
dpc-1863	198	1	he	he	PRON
dpc-1863	198	2	k	k	NOUN
dpc-1863	198	3	,	,	PUNCT
dpc-1863	198	4	gkioxari	gkioxari	VERB
dpc-1863	198	5	g	g	NOUN
dpc-1863	198	6	,	,	PUNCT
dpc-1863	198	7	dollar	dollar	NOUN
dpc-1863	198	8	p	p	NOUN
dpc-1863	198	9	,	,	PUNCT
dpc-1863	198	10	girshick	girshick	PROPN
dpc-1863	198	11	r.	r.	PROPN
dpc-1863	198	12	mask	mask	PROPN
dpc-1863	198	13	r	r	PROPN
dpc-1863	198	14	-	-	PUNCT
dpc-1863	198	15	cnn	cnn	PROPN
dpc-1863	198	16	.	.	PROPN
dpc-1863	199	1	2017	2017	NUM
dpc-1863	199	2	ieee	ieee	PROPN
dpc-1863	199	3	international	international	ADJ
dpc-1863	199	4	conference	conference	NOUN
dpc-1863	199	5	on	on	ADP
dpc-1863	199	6	computer	computer	NOUN
dpc-1863	199	7	vision	vision	NOUN
dpc-1863	199	8	(	(	PUNCT
dpc-1863	199	9	iccv	iccv	PROPN
dpc-1863	199	10	)	)	PUNCT
dpc-1863	199	11	.	.	PUNCT
dpc-1863	200	1	2017:2980	2017:2980	X
dpc-1863	200	2	-	-	SYM
dpc-1863	200	3	2988	2988	NUM
dpc-1863	200	4	.	.	PUNCT
dpc-1863	201	1	doi	doi	NOUN
dpc-1863	201	2	:	:	PUNCT
dpc-1863	201	3	10.1109	10.1109	NUM
dpc-1863	201	4	/	/	SYM
dpc-1863	201	5	iccv.2017.322	iccv.2017.322	PROPN
dpc-1863	201	6	.	.	PROPN
dpc-1863	201	7	15	15	NUM
dpc-1863	201	8	.	.	PUNCT
dpc-1863	202	1	kaur	kaur	PROPN
dpc-1863	202	2	r	r	PROPN
dpc-1863	202	3	,	,	PUNCT
dpc-1863	202	4	leander	leander	PROPN
dpc-1863	202	5	r	r	PROPN
dpc-1863	202	6	,	,	PUNCT
dpc-1863	202	7	mishra	mishra	PROPN
dpc-1863	202	8	nk	nk	PROPN
dpc-1863	202	9	,	,	PUNCT
dpc-1863	202	10	et	et	PROPN
dpc-1863	202	11	al	al	PROPN
dpc-1863	202	12	.	.	PROPN
dpc-1863	202	13	thresholding	thresholde	VERB
dpc-1863	202	14	methods	method	NOUN
dpc-1863	202	15	for	for	ADP
dpc-1863	202	16	lesion	lesion	NOUN
dpc-1863	202	17	segmentation	segmentation	NOUN
dpc-1863	202	18	of	of	ADP
dpc-1863	202	19	basal	basal	ADJ
dpc-1863	202	20	cell	cell	NOUN
dpc-1863	202	21	carcinoma	carcinoma	NOUN
dpc-1863	202	22	in	in	ADP
dpc-1863	202	23	dermoscopy	dermoscopy	NOUN
dpc-1863	202	24	images	image	NOUN
dpc-1863	202	25	.	.	PUNCT
dpc-1863	203	1	skin	skin	NOUN
dpc-1863	203	2	res	re	NOUN
dpc-1863	203	3	technol	technol	NOUN
dpc-1863	203	4	.	.	PUNCT
dpc-1863	204	1	2017;23(3):416	2017;23(3):416	NUM
dpc-1863	204	2	-	-	PUNCT
dpc-1863	204	3	428	428	NUM
dpc-1863	204	4	.	.	PUNCT
dpc-1863	205	1	doi	doi	NOUN
dpc-1863	205	2	:	:	PUNCT
dpc-1863	205	3	10.1111/	10.1111/	NUM
dpc-1863	205	4	srt.12352	srt.12352	NOUN
dpc-1863	205	5	.	.	PUNCT
dpc-1863	206	1	pmid	pmid	NOUN
dpc-1863	206	2	:	:	PUNCT
dpc-1863	206	3	27892649	27892649	NUM
dpc-1863	206	4	.	.	PUNCT
dpc-1863	207	1	16	16	NUM
dpc-1863	207	2	.	.	PUNCT
dpc-1863	208	1	mishra	mishra	PROPN
dpc-1863	208	2	nk	nk	PROPN
dpc-1863	208	3	,	,	PUNCT
dpc-1863	208	4	kaur	kaur	PROPN
dpc-1863	208	5	r	r	NOUN
dpc-1863	208	6	,	,	PUNCT
dpc-1863	208	7	kasmi	kasmi	NOUN
dpc-1863	208	8	r	r	NOUN
dpc-1863	208	9	,	,	PUNCT
dpc-1863	208	10	et	et	PROPN
dpc-1863	208	11	al	al	PROPN
dpc-1863	208	12	.	.	PROPN
dpc-1863	209	1	automatic	automatic	ADJ
dpc-1863	209	2	lesion	lesion	NOUN
dpc-1863	209	3	border	border	NOUN
dpc-1863	209	4	selection	selection	NOUN
dpc-1863	209	5	in	in	ADP
dpc-1863	209	6	dermoscopy	dermoscopy	NOUN
dpc-1863	209	7	images	image	NOUN
dpc-1863	209	8	using	use	VERB
dpc-1863	209	9	morphology	morphology	NOUN
dpc-1863	209	10	and	and	CCONJ
dpc-1863	209	11	color	color	NOUN
dpc-1863	209	12	features	feature	NOUN
dpc-1863	209	13	.	.	PUNCT
dpc-1863	210	1	skin	skin	NOUN
dpc-1863	210	2	res	re	NOUN
dpc-1863	210	3	technol	technol	NOUN
dpc-1863	210	4	.	.	PUNCT
dpc-1863	211	1	2019;25(4):544	2019;25(4):544	PROPN
dpc-1863	211	2	-	-	SYM
dpc-1863	211	3	552	552	NUM
dpc-1863	211	4	.	.	PUNCT
dpc-1863	211	5	doi	doi	NOUN
dpc-1863	211	6	:	:	PUNCT
dpc-1863	211	7	10.1111/	10.1111/	NUM
dpc-1863	211	8	srt.12685	srt.12685	PROPN
dpc-1863	211	9	.	.	PUNCT
dpc-1863	212	1	pmid	pmid	NOUN
dpc-1863	212	2	:	:	PUNCT
dpc-1863	212	3	30868667	30868667	NUM
dpc-1863	212	4	.	.	PUNCT
dpc-1863	213	1	pmcid	pmcid	ADJ
dpc-1863	213	2	:	:	PUNCT
dpc-1863	213	3	pmc7173402	pmc7173402	NOUN
dpc-1863	213	4	.	.	PUNCT
dpc-1863	214	1	17	17	NUM
dpc-1863	214	2	.	.	PUNCT
dpc-1863	214	3	han	han	PROPN
dpc-1863	214	4	ss	ss	PROPN
dpc-1863	214	5	,	,	PUNCT
dpc-1863	214	6	moon	moon	NOUN
dpc-1863	214	7	ij	ij	NOUN
dpc-1863	214	8	,	,	PUNCT
dpc-1863	214	9	lim	lim	PROPN
dpc-1863	214	10	w	w	PROPN
dpc-1863	214	11	,	,	PUNCT
dpc-1863	214	12	et	et	PROPN
dpc-1863	214	13	al	al	PROPN
dpc-1863	214	14	.	.	PROPN
dpc-1863	214	15	keratinocytic	keratinocytic	ADJ
dpc-1863	214	16	skin	skin	NOUN
dpc-1863	214	17	cancer	cancer	NOUN
dpc-1863	214	18	detection	detection	NOUN
dpc-1863	214	19	on	on	ADP
dpc-1863	214	20	the	the	DET
dpc-1863	214	21	face	face	NOUN
dpc-1863	214	22	using	use	VERB
dpc-1863	214	23	region	region	NOUN
dpc-1863	214	24	-	-	PUNCT
dpc-1863	214	25	based	base	VERB
dpc-1863	214	26	convolutional	convolutional	ADJ
dpc-1863	214	27	neural	neural	ADJ
dpc-1863	214	28	network	network	NOUN
dpc-1863	214	29	.	.	PUNCT
dpc-1863	215	1	jama	jama	PROPN
dpc-1863	215	2	dermatol	dermatol	NOUN
dpc-1863	215	3	.	.	PUNCT
dpc-1863	216	1	2020;156(1):29	2020;156(1):29	NUM
dpc-1863	216	2	-	-	PUNCT
dpc-1863	216	3	37	37	NUM
dpc-1863	216	4	.	.	PUNCT
dpc-1863	217	1	doi	doi	NOUN
dpc-1863	217	2	:	:	PUNCT
dpc-1863	217	3	10.1001	10.1001	NUM
dpc-1863	217	4	/	/	SYM
dpc-1863	217	5	jamadermatol.2019.3807	jamadermatol.2019.3807	PROPN
dpc-1863	217	6	.	.	PUNCT
dpc-1863	218	1	pmid	pmid	NOUN
dpc-1863	218	2	:	:	PUNCT
dpc-1863	218	3	31799995	31799995	NUM
dpc-1863	218	4	.	.	PUNCT
dpc-1863	219	1	pmcid	pmcid	NOUN
dpc-1863	219	2	:	:	PUNCT
dpc-1863	219	3	pmc6902187	pmc6902187	NOUN
dpc-1863	219	4	.	.	PUNCT
dpc-1863	220	1	18	18	NUM
dpc-1863	220	2	.	.	PUNCT
dpc-1863	221	1	lee	lee	PROPN
dpc-1863	221	2	t	t	PROPN
dpc-1863	221	3	,	,	PUNCT
dpc-1863	221	4	ng	ng	PROPN
dpc-1863	221	5	v	v	PROPN
dpc-1863	221	6	,	,	PUNCT
dpc-1863	221	7	gallagher	gallagher	PROPN
dpc-1863	221	8	r	r	PROPN
dpc-1863	221	9	,	,	PUNCT
dpc-1863	221	10	coldman	coldman	NOUN
dpc-1863	221	11	a	a	PRON
dpc-1863	221	12	,	,	PUNCT
dpc-1863	221	13	mclean	mclean	PROPN
dpc-1863	221	14	d.	d.	PROPN
dpc-1863	221	15	dullrazor	dullrazor	PROPN
dpc-1863	221	16	:	:	PUNCT
dpc-1863	221	17	a	a	DET
dpc-1863	221	18	software	software	NOUN
dpc-1863	221	19	approach	approach	NOUN
dpc-1863	221	20	to	to	ADP
dpc-1863	221	21	hair	hair	NOUN
dpc-1863	221	22	removal	removal	NOUN
dpc-1863	221	23	from	from	ADP
dpc-1863	221	24	images	image	NOUN
dpc-1863	221	25	.	.	PUNCT
dpc-1863	222	1	original	original	ADJ
dpc-1863	222	2	article	article	NOUN
dpc-1863	222	3	|	|	ADV
dpc-1863	222	4	dermatol	dermatol	ADJ
dpc-1863	222	5	pract	pract	NOUN
dpc-1863	222	6	concept	concept	NOUN
dpc-1863	222	7	.	.	PUNCT
dpc-1863	223	1	2022;12(3):e2022126	2022;12(3):e2022126	NUM
dpc-1863	223	2	7	7	NUM
dpc-1863	223	3	comput	comput	NOUN
dpc-1863	223	4	biol	biol	PROPN
dpc-1863	223	5	med	me	VERB
dpc-1863	223	6	.	.	PUNCT
dpc-1863	224	1	1997;27(6):533	1997;27(6):533	NUM
dpc-1863	224	2	-	-	SYM
dpc-1863	224	3	543	543	NUM
dpc-1863	224	4	.	.	PUNCT
dpc-1863	225	1	doi	doi	NOUN
dpc-1863	225	2	:	:	PUNCT
dpc-1863	225	3	10.1016	10.1016	NUM
dpc-1863	225	4	/	/	SYM
dpc-1863	225	5	s00104825(97)00020	s00104825(97)00020	PROPN
dpc-1863	225	6	-	-	PUNCT
dpc-1863	225	7	6	6	NUM
dpc-1863	225	8	.	.	PUNCT
dpc-1863	226	1	pmid	pmid	NOUN
dpc-1863	226	2	:	:	PUNCT
dpc-1863	226	3	9437554	9437554	NUM
dpc-1863	226	4	.	.	PUNCT
dpc-1863	227	1	19	19	NUM
dpc-1863	227	2	.	.	PUNCT
dpc-1863	227	3	tschandl	tschandl	PROPN
dpc-1863	227	4	p	p	NOUN
dpc-1863	227	5	,	,	PUNCT
dpc-1863	227	6	rosendahl	rosendahl	NOUN
dpc-1863	227	7	c	c	NOUN
dpc-1863	227	8	,	,	PUNCT
dpc-1863	227	9	kittler	kittler	PROPN
dpc-1863	227	10	h.	h.	PROPN
dpc-1863	228	1	the	the	DET
dpc-1863	228	2	ham10000	ham10000	PROPN
dpc-1863	228	3	dataset	dataset	PROPN
dpc-1863	228	4	,	,	PUNCT
dpc-1863	228	5	a	a	DET
dpc-1863	228	6	large	large	ADJ
dpc-1863	228	7	collection	collection	NOUN
dpc-1863	228	8	of	of	ADP
dpc-1863	228	9	multi	multi	ADJ
dpc-1863	228	10	-	-	ADJ
dpc-1863	228	11	source	source	NOUN
dpc-1863	228	12	dermatoscopic	dermatoscopic	NOUN
dpc-1863	228	13	images	image	NOUN
dpc-1863	228	14	of	of	ADP
dpc-1863	228	15	common	common	ADJ
dpc-1863	228	16	pigmented	pigmented	ADJ
dpc-1863	228	17	skin	skin	NOUN
dpc-1863	228	18	lesions	lesion	NOUN
dpc-1863	228	19	.	.	PUNCT
dpc-1863	229	1	sci	sci	PROPN
dpc-1863	229	2	data	data	PROPN
dpc-1863	229	3	.	.	PUNCT
dpc-1863	230	1	2018;5:180161	2018;5:180161	VERB
dpc-1863	230	2	.	.	PUNCT
dpc-1863	231	1	doi	doi	NOUN
dpc-1863	231	2	:	:	PUNCT
dpc-1863	231	3	10.1038/	10.1038/	NUM
dpc-1863	231	4	sdata.2018.161	sdata.2018.161	PROPN
dpc-1863	231	5	.	.	PUNCT
dpc-1863	232	1	pmid	pmid	NOUN
dpc-1863	232	2	:	:	PUNCT
dpc-1863	232	3	30106392	30106392	NUM
dpc-1863	232	4	.	.	PUNCT
dpc-1863	233	1	pmcid	pmcid	NOUN
dpc-1863	233	2	:	:	PUNCT
dpc-1863	233	3	pmc6091241	pmc6091241	NOUN
dpc-1863	233	4	.	.	PUNCT
dpc-1863	234	1	20	20	NUM
dpc-1863	234	2	.	.	PUNCT
dpc-1863	235	1	mendonça	mendonça	PROPN
dpc-1863	235	2	t	t	PROPN
dpc-1863	235	3	,	,	PUNCT
dpc-1863	235	4	ferreira	ferreira	PROPN
dpc-1863	235	5	pm	pm	PROPN
dpc-1863	235	6	,	,	PUNCT
dpc-1863	235	7	marques	marques	PROPN
dpc-1863	235	8	js	js	PROPN
dpc-1863	235	9	,	,	PUNCT
dpc-1863	235	10	marcal	marcal	ADJ
dpc-1863	235	11	ars	ar	NOUN
dpc-1863	235	12	,	,	PUNCT
dpc-1863	235	13	rozeira	rozeira	PROPN
dpc-1863	235	14	j.	j.	PROPN
dpc-1863	235	15	ph2	ph2	PROPN
dpc-1863	235	16	a	a	DET
dpc-1863	235	17	dermoscopic	dermoscopic	ADJ
dpc-1863	235	18	image	image	NOUN
dpc-1863	235	19	database	database	NOUN
dpc-1863	235	20	for	for	ADP
dpc-1863	235	21	research	research	NOUN
dpc-1863	235	22	and	and	CCONJ
dpc-1863	235	23	benchmarking	benchmarking	NOUN
dpc-1863	235	24	.	.	PUNCT
dpc-1863	236	1	in	in	ADP
dpc-1863	236	2	:	:	PUNCT
dpc-1863	236	3	2013	2013	NUM
dpc-1863	236	4	35th	35th	ADJ
dpc-1863	236	5	annual	annual	ADJ
dpc-1863	236	6	international	international	ADJ
dpc-1863	236	7	conference	conference	NOUN
dpc-1863	236	8	of	of	ADP
dpc-1863	236	9	the	the	DET
dpc-1863	236	10	ieee	ieee	NOUN
dpc-1863	236	11	engineering	engineering	NOUN
dpc-1863	236	12	in	in	ADP
dpc-1863	236	13	medicine	medicine	NOUN
dpc-1863	236	14	and	and	CCONJ
dpc-1863	236	15	biology	biology	NOUN
dpc-1863	236	16	society	society	NOUN
dpc-1863	236	17	(	(	PUNCT
dpc-1863	236	18	embc	embc	PROPN
dpc-1863	236	19	)	)	PUNCT
dpc-1863	236	20	.	.	PUNCT
dpc-1863	236	21	;	;	PUNCT
dpc-1863	236	22	2013:5437	2013:5437	NUM
dpc-1863	236	23	-	-	SYM
dpc-1863	236	24	5440	5440	NUM
dpc-1863	236	25	.	.	PUNCT
dpc-1863	237	1	doi	doi	NOUN
dpc-1863	237	2	:	:	PUNCT
dpc-1863	237	3	10.1109	10.1109	NUM
dpc-1863	237	4	/	/	SYM
dpc-1863	237	5	embc.2013.6610779	embc.2013.6610779	PROPN
dpc-1863	237	6	.	.	PUNCT
dpc-1863	238	1	pmid	pmid	PROPN
dpc-1863	238	2	:	:	PUNCT
dpc-1863	238	3	24110966	24110966	NUM
dpc-1863	238	4	.	.	PUNCT
dpc-1863	239	1	21	21	NUM
dpc-1863	239	2	.	.	PUNCT
dpc-1863	239	3	kawahara	kawahara	PROPN
dpc-1863	239	4	j	j	PROPN
dpc-1863	239	5	,	,	PUNCT
dpc-1863	239	6	daneshvar	daneshvar	NOUN
dpc-1863	239	7	s	s	PROPN
dpc-1863	239	8	,	,	PUNCT
dpc-1863	239	9	argenziano	argenziano	ADJ
dpc-1863	239	10	g	g	NOUN
dpc-1863	239	11	,	,	PUNCT
dpc-1863	239	12	hamarneh	hamarneh	PROPN
dpc-1863	239	13	g.	g.	PROPN
dpc-1863	239	14	seven	seven	NUM
dpc-1863	239	15	-	-	PUNCT
dpc-1863	239	16	point	point	NOUN
dpc-1863	239	17	checklist	checklist	NOUN
dpc-1863	239	18	and	and	CCONJ
dpc-1863	239	19	skin	skin	NOUN
dpc-1863	239	20	lesion	lesion	NOUN
dpc-1863	239	21	classification	classification	NOUN
dpc-1863	239	22	using	use	VERB
dpc-1863	239	23	multitask	multitask	ADJ
dpc-1863	239	24	multimodal	multimodal	ADJ
dpc-1863	239	25	neural	neural	ADJ
dpc-1863	239	26	nets	net	NOUN
dpc-1863	239	27	.	.	PUNCT
dpc-1863	240	1	ieee	ieee	PROPN
dpc-1863	240	2	journal	journal	PROPN
dpc-1863	240	3	of	of	ADP
dpc-1863	240	4	biomedical	biomedical	ADJ
dpc-1863	240	5	and	and	CCONJ
dpc-1863	240	6	health	health	NOUN
dpc-1863	240	7	informatics	informatic	NOUN
dpc-1863	240	8	.	.	PUNCT
dpc-1863	241	1	2019;23(2):538	2019;23(2):538	NUM
dpc-1863	241	2	-	-	SYM
dpc-1863	241	3	546	546	NUM
dpc-1863	241	4	.	.	PUNCT
dpc-1863	242	1	doi	doi	NOUN
dpc-1863	242	2	:	:	PUNCT
dpc-1863	242	3	10.1109/	10.1109/	NUM
dpc-1863	242	4	jbhi.2018.2824327	jbhi.2018.2824327	NOUN
dpc-1863	242	5	.	.	PUNCT
dpc-1863	243	1	pmid	pmid	NOUN
dpc-1863	243	2	:	:	PUNCT
dpc-1863	243	3	29993994	29993994	NUM
dpc-1863	243	4	.	.	PUNCT
dpc-1863	244	1	22	22	NUM
dpc-1863	244	2	.	.	PUNCT
dpc-1863	245	1	paszke	paszke	NOUN
dpc-1863	245	2	a	a	DET
dpc-1863	245	3	,	,	PUNCT
dpc-1863	245	4	gross	gross	ADJ
dpc-1863	245	5	s	s	NOUN
dpc-1863	245	6	,	,	PUNCT
dpc-1863	245	7	massa	massa	PROPN
dpc-1863	245	8	f	f	PROPN
dpc-1863	245	9	,	,	PUNCT
dpc-1863	245	10	et	et	PROPN
dpc-1863	245	11	al	al	PROPN
dpc-1863	245	12	.	.	PROPN
dpc-1863	245	13	pytorch	pytorch	PROPN
dpc-1863	245	14	:	:	PUNCT
dpc-1863	245	15	an	an	DET
dpc-1863	245	16	imperative	imperative	ADJ
dpc-1863	245	17	style	style	NOUN
dpc-1863	245	18	,	,	PUNCT
dpc-1863	245	19	high	high	ADJ
dpc-1863	245	20	-	-	PUNCT
dpc-1863	245	21	performance	performance	NOUN
dpc-1863	245	22	deep	deep	ADJ
dpc-1863	245	23	learning	learning	NOUN
dpc-1863	245	24	library	library	NOUN
dpc-1863	245	25	.	.	PUNCT
dpc-1863	246	1	in	in	ADP
dpc-1863	246	2	:	:	PUNCT
dpc-1863	246	3	wallach	wallach	PROPN
dpc-1863	246	4	h	h	PROPN
dpc-1863	246	5	,	,	PUNCT
dpc-1863	246	6	larochelle	larochelle	PROPN
dpc-1863	246	7	h	h	NOUN
dpc-1863	246	8	,	,	PUNCT
dpc-1863	246	9	beygelzimer	beygelzimer	NOUN
dpc-1863	246	10	a	a	X
dpc-1863	246	11	,	,	PUNCT
dpc-1863	246	12	d	d	NOUN
dpc-1863	246	13	’	'	PUNCT
dpc-1863	246	14	alché	alché	ADJ
dpc-1863	246	15	-	-	PUNCT
dpc-1863	246	16	buc	buc	NOUN
dpc-1863	246	17	f	f	NOUN
dpc-1863	246	18	,	,	PUNCT
dpc-1863	246	19	fox	fox	PROPN
dpc-1863	246	20	e	e	PROPN
dpc-1863	246	21	,	,	PUNCT
dpc-1863	246	22	garnett	garnett	PROPN
dpc-1863	246	23	r	r	PROPN
dpc-1863	246	24	,	,	PUNCT
dpc-1863	246	25	eds	ed	NOUN
dpc-1863	246	26	.	.	PUNCT
dpc-1863	246	27	advances	advance	NOUN
dpc-1863	246	28	in	in	ADP
dpc-1863	246	29	neural	neural	ADJ
dpc-1863	246	30	information	information	NOUN
dpc-1863	246	31	processing	processing	NOUN
dpc-1863	246	32	systems	system	NOUN
dpc-1863	246	33	32	32	NUM
dpc-1863	246	34	.	.	PUNCT
dpc-1863	247	1	curran	curran	PROPN
dpc-1863	247	2	associates	associates	PROPN
dpc-1863	247	3	,	,	PUNCT
dpc-1863	247	4	inc	inc	PROPN
dpc-1863	247	5	.	.	PROPN
dpc-1863	247	6	;	;	PUNCT
dpc-1863	247	7	2019:8026	2019:8026	NUM
dpc-1863	247	8	-	-	SYM
dpc-1863	247	9	8037	8037	NUM
dpc-1863	247	10	.	.	PUNCT
dpc-1863	248	1	available	available	ADJ
dpc-1863	248	2	from	from	ADP
dpc-1863	248	3	:	:	PUNCT
dpc-1863	248	4	https://papers.nips.cc/paper/2019/hash/bdbca288fee7f92f2bfa9f7012727740-abstract.html	https://papers.nips.cc/paper/2019/hash/bdbca288fee7f92f2bfa9f7012727740-abstract.html	PROPN
dpc-1863	248	5	23	23	NUM
dpc-1863	248	6	.	.	PUNCT
dpc-1863	249	1	pedregosa	pedregosa	PROPN
dpc-1863	249	2	f	f	PROPN
dpc-1863	249	3	,	,	PUNCT
dpc-1863	249	4	varoquaux	varoquaux	VERB
dpc-1863	249	5	g	g	NOUN
dpc-1863	249	6	,	,	PUNCT
dpc-1863	249	7	gramfort	gramfort	VERB
dpc-1863	249	8	a	a	PRON
dpc-1863	249	9	,	,	PUNCT
dpc-1863	249	10	et	et	PROPN
dpc-1863	249	11	al	al	PROPN
dpc-1863	249	12	.	.	PROPN
dpc-1863	249	13	scikit	scikit	PROPN
dpc-1863	249	14	-	-	PUNCT
dpc-1863	249	15	learn	learn	VERB
dpc-1863	249	16	:	:	PUNCT
dpc-1863	249	17	machine	machine	NOUN
dpc-1863	249	18	learning	learning	NOUN
dpc-1863	249	19	in	in	ADP
dpc-1863	249	20	python	python	PROPN
dpc-1863	249	21	.	.	PUNCT
dpc-1863	250	1	the	the	DET
dpc-1863	250	2	journal	journal	NOUN
dpc-1863	250	3	of	of	ADP
dpc-1863	250	4	machine	machine	NOUN
dpc-1863	250	5	learning	learn	VERB
dpc-1863	250	6	research	research	NOUN
dpc-1863	250	7	.	.	PUNCT
dpc-1863	251	1	2011;12:2825	2011;12:2825	PROPN
dpc-1863	251	2	-	-	PUNCT
dpc-1863	251	3	2830	2830	NUM
dpc-1863	251	4	.	.	PUNCT
dpc-1863	252	1	available	available	ADJ
dpc-1863	252	2	from	from	ADP
dpc-1863	252	3	:	:	PUNCT
dpc-1863	252	4	https://www.jmlr	https://www.jmlr	NOUN
dpc-1863	252	5	.	.	PUNCT
dpc-1863	252	6	org	org	ADJ
dpc-1863	252	7	/	/	SYM
dpc-1863	252	8	papers	paper	NOUN
dpc-1863	252	9	/	/	SYM
dpc-1863	252	10	volume12	volume12	NOUN
dpc-1863	252	11	/	/	PUNCT
dpc-1863	252	12	pedregosa11a	pedregosa11a	PROPN
dpc-1863	252	13	/	/	SYM
dpc-1863	252	14	pedregosa11a.pdf	pedregosa11a.pdf	NOUN
dpc-1863	252	15	24	24	NUM
dpc-1863	252	16	.	.	PUNCT
dpc-1863	253	1	seabold	seabold	PROPN
dpc-1863	253	2	s	s	PROPN
dpc-1863	253	3	,	,	PUNCT
dpc-1863	253	4	perktold	perktold	PROPN
dpc-1863	253	5	j.	j.	PROPN
dpc-1863	253	6	statsmodels	statsmodels	PROPN
dpc-1863	253	7	:	:	PUNCT
dpc-1863	254	1	econometric	econometric	ADJ
dpc-1863	254	2	and	and	CCONJ
dpc-1863	254	3	statistical	statistical	ADJ
dpc-1863	254	4	modeling	modeling	NOUN
dpc-1863	254	5	with	with	ADP
dpc-1863	254	6	python	python	PROPN
dpc-1863	254	7	.	.	PUNCT
dpc-1863	255	1	in	in	ADP
dpc-1863	255	2	:	:	PUNCT
dpc-1863	255	3	proceedings	proceeding	NOUN
dpc-1863	255	4	of	of	ADP
dpc-1863	255	5	the	the	DET
dpc-1863	255	6	9th	9th	ADJ
dpc-1863	255	7	python	python	NOUN
dpc-1863	255	8	in	in	ADP
dpc-1863	255	9	science	science	NOUN
dpc-1863	255	10	conference	conference	NOUN
dpc-1863	255	11	.	.	PUNCT
dpc-1863	256	1	vol	vol	NOUN
dpc-1863	256	2	57	57	NUM
dpc-1863	256	3	.	.	PUNCT
dpc-1863	257	1	austin	austin	PROPN
dpc-1863	257	2	,	,	PUNCT
dpc-1863	257	3	tx	tx	PROPN
dpc-1863	257	4	;	;	PUNCT
dpc-1863	257	5	2010:61	2010:61	NUM
dpc-1863	257	6	.	.	PUNCT
dpc-1863	258	1	available	available	ADJ
dpc-1863	258	2	from	from	ADP
dpc-1863	258	3	:	:	PUNCT
dpc-1863	258	4	https://	https://	PROPN
dpc-1863	258	5	conference.scipy.org/proceedings/scipy2010/pdfs/seabold.pdf	conference.scipy.org/proceedings/scipy2010/pdfs/seabold.pdf	PROPN
dpc-1863	258	6	25	25	NUM
dpc-1863	258	7	.	.	PUNCT
dpc-1863	259	1	maron	maron	PROPN
dpc-1863	259	2	rc	rc	PROPN
dpc-1863	259	3	,	,	PUNCT
dpc-1863	259	4	haggenmüller	haggenmüller	NOUN
dpc-1863	259	5	s	s	PROPN
dpc-1863	259	6	,	,	PUNCT
dpc-1863	259	7	von	von	PROPN
dpc-1863	259	8	kalle	kalle	PROPN
dpc-1863	259	9	c	c	PROPN
dpc-1863	259	10	,	,	PUNCT
dpc-1863	259	11	et	et	PROPN
dpc-1863	259	12	al	al	PROPN
dpc-1863	259	13	.	.	PUNCT
dpc-1863	260	1	robustness	robustness	NOUN
dpc-1863	260	2	of	of	ADP
dpc-1863	260	3	convolutional	convolutional	ADJ
dpc-1863	260	4	neural	neural	ADJ
dpc-1863	260	5	networks	network	NOUN
dpc-1863	260	6	in	in	ADP
dpc-1863	260	7	recognition	recognition	NOUN
dpc-1863	260	8	of	of	ADP
dpc-1863	260	9	pigmented	pigmented	ADJ
dpc-1863	260	10	skin	skin	NOUN
dpc-1863	260	11	lesions	lesion	NOUN
dpc-1863	260	12	.	.	PUNCT
dpc-1863	261	1	eur	eur	PROPN
dpc-1863	261	2	j	j	PROPN
dpc-1863	261	3	cancer	cancer	NOUN
dpc-1863	261	4	.	.	PUNCT
dpc-1863	262	1	2021;145:81	2021;145:81	NUM
dpc-1863	262	2	-	-	SYM
dpc-1863	262	3	91	91	NUM
dpc-1863	262	4	.	.	PUNCT
dpc-1863	263	1	doi	doi	NOUN
dpc-1863	263	2	:	:	PUNCT
dpc-1863	263	3	10.1016	10.1016	NUM
dpc-1863	263	4	/	/	SYM
dpc-1863	263	5	j.	j.	PROPN
dpc-1863	263	6	ejca.2020.11.020	ejca.2020.11.020	PROPN
dpc-1863	263	7	.	.	PUNCT
dpc-1863	264	1	pmid	pmid	NOUN
dpc-1863	264	2	:	:	PUNCT
dpc-1863	264	3	33423009	33423009	NUM
dpc-1863	264	4	.	.	PUNCT
dpc-1863	265	1	26	26	NUM
dpc-1863	265	2	.	.	PUNCT
dpc-1863	266	1	aggarwal	aggarwal	PROPN
dpc-1863	266	2	slp	slp	PROPN
dpc-1863	266	3	.	.	PUNCT
dpc-1863	267	1	data	datum	NOUN
dpc-1863	267	2	augmentation	augmentation	NOUN
dpc-1863	267	3	in	in	ADP
dpc-1863	267	4	dermatology	dermatology	NOUN
dpc-1863	267	5	image	image	NOUN
dpc-1863	267	6	recognition	recognition	NOUN
dpc-1863	267	7	using	use	VERB
dpc-1863	267	8	machine	machine	NOUN
dpc-1863	267	9	learning	learning	NOUN
dpc-1863	267	10	.	.	PUNCT
dpc-1863	268	1	skin	skin	NOUN
dpc-1863	268	2	res	re	NOUN
dpc-1863	268	3	technol	technol	NOUN
dpc-1863	268	4	.	.	PUNCT
dpc-1863	269	1	2019;25(6):815820	2019;25(6):815820	NUM
dpc-1863	269	2	.	.	PUNCT
dpc-1863	270	1	doi	doi	NOUN
dpc-1863	270	2	:	:	PUNCT
dpc-1863	270	3	10.1111	10.1111	NUM
dpc-1863	270	4	/	/	SYM
dpc-1863	270	5	srt.12726	srt.12726	PROPN
dpc-1863	270	6	.	.	NOUN
dpc-1863	270	7	pmid	pmid	NOUN
dpc-1863	270	8	:	:	PUNCT
dpc-1863	270	9	31140653	31140653	NUM
dpc-1863	270	10	.	.	PUNCT
dpc-1863	271	1	27	27	NUM
dpc-1863	271	2	.	.	PUNCT
dpc-1863	271	3	winkler	winkler	PROPN
dpc-1863	271	4	jk	jk	PROPN
dpc-1863	271	5	,	,	PUNCT
dpc-1863	271	6	sies	sies	PROPN
dpc-1863	271	7	k	k	PROPN
dpc-1863	271	8	,	,	PUNCT
dpc-1863	271	9	fink	fink	PROPN
dpc-1863	271	10	c	c	PROPN
dpc-1863	271	11	,	,	PUNCT
dpc-1863	271	12	et	et	PROPN
dpc-1863	271	13	al	al	PROPN
dpc-1863	271	14	.	.	PROPN
dpc-1863	271	15	association	association	PROPN
dpc-1863	271	16	between	between	ADP
dpc-1863	271	17	different	different	ADJ
dpc-1863	271	18	scale	scale	NOUN
dpc-1863	271	19	bars	bar	NOUN
dpc-1863	271	20	in	in	ADP
dpc-1863	271	21	dermoscopic	dermoscopic	ADJ
dpc-1863	271	22	images	image	NOUN
dpc-1863	271	23	and	and	CCONJ
dpc-1863	271	24	diagnostic	diagnostic	ADJ
dpc-1863	271	25	performance	performance	NOUN
dpc-1863	271	26	of	of	ADP
dpc-1863	271	27	a	a	DET
dpc-1863	271	28	market	market	NOUN
dpc-1863	271	29	-	-	PUNCT
dpc-1863	271	30	approved	approve	VERB
dpc-1863	271	31	deep	deep	ADJ
dpc-1863	271	32	learning	learn	VERB
dpc-1863	271	33	convolutional	convolutional	ADJ
dpc-1863	271	34	neural	neural	ADJ
dpc-1863	271	35	network	network	NOUN
dpc-1863	271	36	for	for	ADP
dpc-1863	271	37	melanoma	melanoma	NOUN
dpc-1863	271	38	recognition	recognition	NOUN
dpc-1863	271	39	.	.	PUNCT
dpc-1863	272	1	eur	eur	PROPN
dpc-1863	272	2	j	j	PROPN
dpc-1863	272	3	cancer	cancer	NOUN
dpc-1863	272	4	.	.	PUNCT
dpc-1863	273	1	2021;145:146	2021;145:146	NUM
dpc-1863	273	2	-	-	SYM
dpc-1863	273	3	154	154	NUM
dpc-1863	273	4	.	.	PUNCT
dpc-1863	274	1	doi	doi	NOUN
dpc-1863	274	2	:	:	PUNCT
dpc-1863	274	3	10.1016	10.1016	NUM
dpc-1863	274	4	/	/	SYM
dpc-1863	274	5	j.ejca.2020.12.010	j.ejca.2020.12.010	PROPN
dpc-1863	274	6	.	.	PUNCT
dpc-1863	275	1	pmid	pmid	NOUN
dpc-1863	275	2	:	:	PUNCT
dpc-1863	275	3	33465706	33465706	NUM
dpc-1863	275	4	.	.	PUNCT
dpc-1863	276	1	28	28	NUM
dpc-1863	276	2	.	.	PUNCT
dpc-1863	277	1	michaelis	michaelis	PROPN
dpc-1863	277	2	c	c	X
dpc-1863	277	3	,	,	PUNCT
dpc-1863	277	4	mitzkus	mitzkus	PROPN
dpc-1863	277	5	b	b	PROPN
dpc-1863	277	6	,	,	PUNCT
dpc-1863	277	7	geirhos	geirhos	PROPN
dpc-1863	277	8	r	r	NOUN
dpc-1863	277	9	,	,	PUNCT
dpc-1863	277	10	et	et	PROPN
dpc-1863	277	11	al	al	PROPN
dpc-1863	277	12	.	.	PUNCT
dpc-1863	278	1	benchmarking	benchmarke	VERB
dpc-1863	278	2	robustness	robustness	NOUN
dpc-1863	278	3	in	in	ADP
dpc-1863	278	4	object	object	NOUN
dpc-1863	278	5	detection	detection	NOUN
dpc-1863	278	6	:	:	PUNCT
dpc-1863	278	7	autonomous	autonomous	ADJ
dpc-1863	278	8	driving	driving	NOUN
dpc-1863	278	9	when	when	SCONJ
dpc-1863	278	10	winter	winter	NOUN
dpc-1863	278	11	is	be	AUX
dpc-1863	278	12	coming	come	VERB
dpc-1863	278	13	.	.	PUNCT
dpc-1863	279	1	arxiv	arxiv	PROPN
dpc-1863	280	1	[	[	X
dpc-1863	280	2	cscv	cscv	NOUN
dpc-1863	280	3	]	]	X
dpc-1863	280	4	.	.	PUNCT
dpc-1863	281	1	published	publish	VERB
dpc-1863	281	2	online	online	ADJ
dpc-1863	281	3	july	july	PROPN
dpc-1863	281	4	17	17	NUM
dpc-1863	281	5	,	,	PUNCT
dpc-1863	281	6	2019	2019	NUM
dpc-1863	281	7	.	.	PUNCT
dpc-1863	282	1	available	available	ADJ
dpc-1863	282	2	from	from	ADP
dpc-1863	282	3	:	:	PUNCT
dpc-1863	282	4	http://arxiv.org/abs/1907.07484	http://arxiv.org/abs/1907.07484	NOUN