Quantitative Analysis of Ultrasound Tissue Diffusion Elastography in The Diagnosis of Benign and Malignant Prostate Lesions Jun Guo1,2, Lei Liang2, Nan Zhou2, De-Yu Li1* Purpose: This study aims to evaluate the value of quantitative analysis of ultrasound real-time tissue diffusion elastography in the diagnosis of benign and malignant prostate lesions. Materials and Methods: From March 2010 to June 2013, 52 patients suspected with prostate cancer based on laboratory or clinical test results and underwent prostate biopsy in our hospital were enrolled into this study. The age of these patients ranged between 45-82 years, with an average age of 67.2 ± 6.8 years. All patients underwent transrectal real-time ultrasound elastography (TRTE) before biopsy. A total of 63 prostate nodules were detected, and the 11 elastic characteristic quantities of these nodules were quantitatively analyzed via tissue diffusion quan- titative analysis. The results of ultrasonography were compared with the results of operation and pathology. Result: Among these 11 characteristic quantities, which include the mean (MEAN) and standard deviation (SD), blue area ratio (AREA%), complexity (COMP), kurtosis (KURT), skewness (SKEW), contrast (CONT), equality (ENT), entropy (IDM), consistency (ASM) and correlation (CORR), except for COMP and CORR, the differences in other nine characteristic quantities between benign and malignant prostatic nodules were statistically signifi- cant (P<0.05). Among these, the AREA% and MEAN had the highest correlation, which were 0.791 and -0.791, respectively. The Youden’s index (sensitivity and specificity) of AREA% in the ROC curves was the highest, the cutoff value was 80.65% for the diagnosis of prostate cancer, sensitivity was 87.9%, and specificity was 96.6%. Conclusion: Quantitative analysis of ultrasound real-time tissue diffusion elastography is helpful in the diagnosis of benign and malignant prostate lesions, provides a relatively accurate evaluation method in clinical practice, and has broad application prospects. Keywords: ultrasound; elastography; prostate cancer; tissue dispersion quantitative analysis; biopsy INTRODUCTION Prostate cancer has become one of the major diseases that threatens the health of elderly men. In recent years, its incidence has increased annually in China. Furthermore, the missed diagnosis rate of digital rectal examination in the diagnosis of cancers is as high as 80%, hence, its value is very limited.(1) Prostate-specif- ic antigen (PSA) detection and transrectal ultrasonog- raphy (TRUS) greatly improves the early diagnosis of prostate cancer, and are significantly better than digital rectal examination.(2) However, 25-45% of prostate cancer patients present with normal PSA levels.(3) At present, the conventional diagnosis of prostate nodules by TRUS is mainly ac- cording to the location, shape, envelope, internal echo and blood flow at the sites of the lesions; which has low sensitivity and specificity.(4) In addition, cell density in prostate cancer is higher than in normal tissues, and this increase in cell density would lead to changes in tissue elasticity.(5) 1School of Biological Science and Medical Engineering, Beihang University, Beijing, 100191, China. 2Department of Ultrasound, Aero Space Central Hospital, Beijing, 100049, China. *Correspondence: School of Biological Science and Medical Engineering, Beihang University, No.37 Xueyuan Road, Haidian District, Beijing,100191, China. Tel: + 86-10-82339093, Fax: + 86-10-82339093, E-mail: lideyu_dn@sina.com. Received November 2017 & Accepted December 2018 Clinical pathology studies have shown that there are significant differences in the elastic coefficients of fat, breast, fibrotic tissue, non-invasive carcinogenesis and invasive cancerous tissue. Therefore, the degree of deformation of different tissues under different ex- ternal forces will be different. The tissue elastography technique is based on this theory and receives the in- formation of the tissue displacement after compression by echo, and displays it by color signal encoding after computer super-high speed processing. Real-time tis- sue elastography (RTE) is by means of exert internal pressure or external pressure on tissues with the probe. Under the conditions of elastic mechanics and biome- chanics, these tissues generate a response such as the distribution of displacement, strain and velocity. Ac- cording to differences in the elastic coefficient of tis- sues, the strains vary. Furthermore, the echo signals be- fore and after compression were collected and analyzed. According to the displacement of different positions in the tissues, the deformation degree is calculated and displayed as images using grayscale or color coding. UROLOGICAL ONCOLOGY Urology Journal/Vol 16 No. 4/ July-August 2019/ pp. 347-351. [DOI: http://dx.doi.org/10.22037/uj.v0i0.4224] Vol 16 No 04 July-August 2019 348 In the present study, with postoperative puncture and pathological results as the gold standard, RTE and the quantitative analysis of tissue diffusion were applied in the differential diagnosis of benign and malignant pros- tate lesions to evaluate its diagnostic value. MATERIALS AND METHODS Research Data This study was approved by the ethical committee. From March 2010 to June 2013, 52 patients suspected with prostate cancer based on laboratory or clinical test results and underwent prostate biopsy in our hospital were enrolled into this study. The age of these patients ranged between 45-82 years, with an average age of 67.2 ± 6.8 years. Exclusion criteria were: patients who received treatment for prostate disease before admis- sion including transurethral resection of the prostate, particle implantation for prostate cancer, and prostate hormone therapy, patients with a prostate size exceed- ing the transrectal real-time ultrasound elastography (TRTE) range, and patients who underwent prostate puncture within one year. All patients underwent ultra- sound-guided biopsy or surgery to obtain pathological results. There are two pathologists who both had ex- perience over ten years and they were unaware of the clinical conditions and ultrasound performance of the patients. All patients agreed to participate in this study and provided a signed informed consent. Instruments and Methods The HiVision 900 color Doppler ultrasound diagnostic apparatus (Tokyo, Japan) was used for this study, with the EUP-V53W probe. The frequency ranged between 4-9 MHz. The patient was positioned in the left later- al position and underwent routine TRUS to detect the nodules and observe the morphology, envelope, inter- nal echo and blood flow, and measure systolic velocity (Vs) and diastolic velocity (Vd) of the blood flow and resistance index (RI) within the lesions. The mode was switched to elastic imaging, and the gray-scale map and elastic graph were simultaneously observed using the dual display function. The probe was adjusted to show the nodule at the center of the screen, pressure was manually exerted on the prostate using a transrectal probe, and elastic imaging was performed by steady small jitter until a stable and repeatable dynamic elastic graph was obtained. If nodules were not found by two-dimensional ultrasound, the mode was switched to elastic imaging after conventional TRUS; and the whole prostate was first observed. Then, the probe was laterally adjusted, and the bilateral internal prostate gland, transitional zone and the outer gland region were observed. The pressure value was displayed as a num- ber within 1-5 in real-time. It was only when the elastic images were continuously displayed and the pressure values were within 4-5 that the test results could be evaluated with valid results. Quantitative Measurement of Tissue Diffusion The image was analyzed using the random elastic tissue diffusion quantitative analysis software. The position of the sampling frame did not exceed the largest rectan- gular area of the scope of the lesion. Measurement was conducted three times, and the mean was calculated. Hence, 11 quantitative parameters of tissue diffusion were obtained, including the mean (MEAN) and stand- ard deviation (SD), blue area ratio (AREA%), com- plexity (COMP), kurtosis (KURT), skewness (SKEW), contrast (CONT), equality (ENT), entropy (IDM), con- sistency (ASM) and correlation (CORR). Statistical Analysis Data were analyzed using statistical software SPSS 12.0. The 11 characteristic quantities of the elastic im- aging of benign and malignant lesions of the prostate were tested for normal distribution. Data in normal distribution were compared using t-test, and data in non-normal distribution were compared using nonpara- metric U-test. P < 0.05 was considered statistically sig- nificant. The correlation between elastic characteristic quantities and pathological grades was analyzed using Spearman correlation analysis. The receiver operating characteristic curve (ROC) was drawn using significant characteristic quantities. Then, the sensitivity and spec- ificity of each characteristic quantity in the diagnosis of prostate cancer was calculated. RESULTS Among the 52 patients in study group, a total of 63 pros- tate lesions were detected by TRTE. It was confirmed by biopsy that 20 lesions were malignant. Among these lesions, 12 nodules had a Gleason score of 2-6 points, two nodules had a Gleason score of 7 points, and six nodules had a Gleason score of 8-10 points. Further- more, 43 nodules were benign lesions. Among these lesions, nine cases were combined with chronic in- flammation of the prostate and one case was combined with grade II of prostatic intraepithelial neoplasm (PIN, Figures 1-3). Furthermore, the PSA levels of these in- cluded patients ranged between 0.38-100 ng/mL, with an average value of 39.9 ng/mL and a median value of 18.7 ng/mL. Among these patients, seven patients had a PSA level of < 4 ng/mL, 13 patients had a PSA level of > 4 ng/mL and <10 ng/mL, and 32 patients had a PSA level of ≥ 10 ng/mL. The 11 characteristic quantities of prostate lesions were determined to be not in the normal distribution. Nonpar- ametric U-test analysis revealed that, except for COMP and CORR, differences in other nine characteristic quantities between benign and malignant prostatic nod- ules were statistically significant (P < 0.05). Correlation analysis revealed that AREA%, IDM, SKEW, KURT and ASM were positively correlated with the malignant Ultrasonic elastography in prostatic diseases-Guo et al. Table 1. Analysis of elastic characteristic quantity of benign and malignant lesions of prostate. Group Number of Lesions MEAN SD AREA% COMP KURT SKEW CONT ENT IDM ASM Optimum 29 86.75 ± 30.49 38.56 ± 12.77 36.24 ± 27.52 25.83 ± 8.97 3.15 ± 1.11 0.65 ± 0.47 29.97 ± 5.57 3.18 ± 0.35 0.26 ± 0.08 0.00 ± 0.00 Malignant 33 18.18 ± 3.16 19.76 ± 12.67 91.09 ± 14.29 25.14 ± 17.64 6.60 ± 4.16 1.57 ± 0.77 24.87 ± 4.33 2.22 ± 0.62 0.57 ± 0.18 0.09 ± 0.14 F value 11.49 6.01 21.401 4.502 18.828 7.397 1.565 4.751 17.514 34.816 P valuea 0.000 0.016 0.000 0.038 0.000 0.009 0.216 0.033 0.000 0.000 a P < 0.05 was considered statistically significant. Urological Oncology 349 degree of lesions; and the correlation coefficients were 0.791, 0.754, 0.581, 0.488 and 0.398, respectively. Fur- thermore, the correlation with AREA% was the highest. MEAN, ENT, SD and CONT were negatively correlat- ed with the malignant degree of lesions, and the correla- tion coefficients were -0.791, -0.721, -0.600 and -0.466, respectively. Furthermore, correlation with MEAN was the highest (Table 1). The ROC for the diagnosis of prostate cancer was drawn using the 11 characteristic quantities of elastic imaging. The Youden’s index (sensitivity and speci- ficity) of AREA% was the highest, the area under the curve (AUC) was 0.961, the diagnostic cutoff value was 80.65%, sensitivity was 87.9%, and specificity was 96.6%. Based on the Youden's index (from high to low), the diagnostic efficiencies of the other eight character- Figure 1. Different types of prostate pathological changes of color doppler ultrasound and ultrasound elasticity imaging. 1a Right low echo between inner and outer gland nodules,Inside, and high resistance arterial spectrum, TRUS malignant likely prompt. 1b The right between the inner and outer gland nodules,TRTE showed lesions subject for the green,with a little blue, score, level, a diagnosis of benign hyperplastic nodule. Confirmed by biopsy for prostate hyperplasia, interstitial small amounts of acute or chronic inflammatory cells infiltration. Figure 2 . Different types of prostate pathological changes of color doppler ultrasound and ultrasound elasticity imaging. 2a The left outer gland low echo nodules, inside, and high resistance arterial spectrum，TRUS malignant likely prompt. 2b The left outer gland nodules is almost covered by blue, blue and lesion area roughly the same，score ， level，a diagnosis of malignant. Confirmed by biopsy for moderately differentiated adenocarcinoma of prostate, Gleason score 3 + 3 = 6 points. Figure 4 . Positive correlation characteristics of benign and malig- nant lesions of the prostate diagnostic value of ROC curve. Figure 3. Different types of prostate pathological changes of color doppler ultrasound and ultrasound elasticity imaging. 3a The left side is no clear boundary between internal and external gland，dif- fuse to reduce，on the right side of the outer gland low echo nod- ules，TRUS prompt malignant. 3b Lesions completely covered by blue, and the blue range is greater than the TRUS detected lesions area, involvement of bilateral internal and external gland, TRTE score ， level, a diagnosis of malignant, Confirmed by biopsy in poorly differentiated adenocarcinoma of prostate, Gleasonscore 5 + 4 = 9 points, both inner and outer gland needle puncture eight right carcinoma tissue. Ultrasonic elastography in prostatic diseases-Guo et al. Vol 16 No 04 July-August 2019 350 istic quantities were MEAN, ASM, IDM, ENT, CONT, SD, SKEW and KURT, respectively; and the cutoff values were 45.15, 0.005, 0.315, 2.695, 16.65, 23.25, 0.93 and 3.75, respectively. Furthermore, the sensitivi- ties were 87.9%, 93.1%, 84.8%, 93.9%, 96.6%, 93.1%, 89.7% and 78.8%, respectively; and the specificities were 96.6%, 90.9%, 93.1%, 82.8%, 78.8%, 81.8%, 69.7% and 79.3%, respectively (Figures 4 and 5). DISCUSSION Since 1991, real-time tissue diffusion elastography was proposed, and was gradually developed and applied. At present, it has gradually become mature in the diagno- sis of breast and thyroid diseases. In the past, the diag- nosis method of prostate elasticity was Kamoi 5-grade scoring.(6) Its principle is to exert internal pressure such as heart beat and respiration or external pressure on tissues with the probe. Under the conditions of elastic mechanics and biomechanics, these tissues generate a response such as the distribution of displacement, strain and velocity. According to differences in the elastic coefficient which is stress/strain of tissues, the strains vary. Furthermore, the echo signals before and after compression were collected and analyzed. According to the displacement of different positions in the tissues, the deformation degree is calculated and displayed as images using grayscale or color coding. In recent years, TRTE transforms the migration amplitude of echo sig- nals before and after compression into real-time color images, and describes the soft, medium hard and hard textures as red, green and blue colors, respectively. The large change in the displacement of tissues after compression is shown as red, the small change in the displacement of tissues after the compression is shown as blue, and the medium change in the displacement of tissues after the compression is shown as green. The elasticity of different tissues is encoded using color, which reflects its relative hardness.(7-9) Kamoi K et al. confirmed that the above methods were helpful in im- proving the diagnostic rate of prostate lesions.(7) On this basis, Hitachi has developed the tissue diffusion quantitative analysis software, which obtains 11 elas- tic characteristic quantities through the measurement of the strain histogram, for the quantitative evaluation of tissue flexibility. The core of this method is the charac- teristic quantity extraction, grading and analysis of the degree of tissue diffusion in real-time tissue elastogra- phy.(10) In this study, this technique was used for the quantitative evaluation of prostate gland lesions. Re- sults revealed that among the 11 elastic characteristic quantities, except for COMP and CORR, differences in other nine characteristic quantities between benign and malignant prostatic nodules were statistically signifi- cant (P < 0.05). Among these, AREA% and MEAN had the highest correlation, the Youden’s index of AREA% was the highest, and the cutoff value in the diagnosis of prostate cancer was 80.65%. In this study, the AREA% was lower than this cut-off value in two cases of malig- nant lesions, and the clinical stages of these two cases were highly differentiated adenocarcinoma. From the pathological perspective, due to the retention of the lu- men, highly differentiated adenocarcinoma is very sim- ilar to normal prostate glands.(11-13) Therefore, in further studies, the relationship between the degree of differ- entiation of prostate cancer and the hardness of cancer tissues should be examined, in order to provide a more accurate preoperative assessment in clinical practice. However, this study had some limitations. First, sample size needs to be enlarged in the future study. Second, the external pressure on tissues with the probe have subjectivity. ACKNOWLEDGEMENT This study was supported by the Foundation of Aero Space Central Hospital. (201124). CONFLICT ON INTEREST The authors have no conflicts of interest to declare. REFERENCES 1. Nygård Y, Haukaas SA, Halvorsen OJ, et al. A positive Real-Time Elastography (RTE) combined with a Prostate Cancer Gene 3 (PCA3) score above 35 convey a high probability of intermediate- or high-risk prostate cancer in patient admitted for primary prostate biopsy. BMC Urol. 2016;16;1:1-8. 2. Heidenreich A, Bastian PJ, Bellmunt J, et al. EAU guidelines on prostate cancer. Part 1: screening, diagnosis, and local treatment with curative intent—update 2013. Euro urol. 2014; 65;1: 124-37. 3. Catalona WJ, Smith DS, Ratliff TL, et al. Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med 1991; 324;17: 1156-61. 4. Walz J, Loch T, Salomon G, Wijkstra H. [Imaging of the prostate][J]. Der Urologe. Ausg. A 2013; 52;4: 490-6. 5. Zhang Q,Yao J,Cai Y, Zhang L, Wu Y, Xiong J, Shi J, Wang Y, Wang Y.Elevated hardness of peripheral gland on real-time elastography is an independent marker for high-risk prostate cancers.La Radiologia medica,2017,122:944-51. 6. Brock M, Eggert T, Palisaar RJ, et al. Multiparametric ultrasound of the prostate: Figure 5. Negative correlation characteristics of benign and malig- nant lesions of the prostate diagnostic value of ROC curve. Ultrasonic elastography in prostatic diseases-Guo et al. adding contrast enhanced ultrasound to real- time elastography to detect histopathologically confirmed cancer. J Urol 2013; 189;1: 93-8. 7. Kamoi K, Okihara K, Ochiai A, et al. The utility of transrectal real-time elastography in the diagnosis of prostate cancer. Ultrasound Med Biol 2008; 34;7: 1025-32. 8. Porsch M, Wendler JJ, Liehr UB, Lux A, Schostak M and Pech M. New aspects in shear-wave elastography of prostate cancer. J Radiol 2015; 15;60: 5. 9. Brock M, Eggert T, Palisaar RJ, et al. Multiparametric ultrasound of the prostate: adding contrast enhanced ultrasound to real- time elastography to detect histopathologically confirmed cancer. J Urol 2013; 189: 93-8. 10. Tatsumi C, Kudo M, Ueshima K, et al. Non- invasive evaluation of hepatic fibrosis for type C chronic hepatitis. Intervirology 2010; 53;1: 76-1. 11. Brock M, Sommerer F, Löppenberg B, et al. Comparison of real-time elastography with grey-scale ultrasonography for detection of organ-confined prostate cancer and extra capsular extension: a prospective analysis using whole mount sections after radical prostatectomy. BJU Int 2011; 108;8: E217-E222. 12. Walz J, Marcy M, Maubon T, et al. Real time elastography in the diagnosis of prostate cancer: comparison of preoperative imaging and histology after radical prostatectomy. Ultrasound Med Biol 2011; 21;13: 925-31. 13. Salomon G, Köllerman J, Thederan I, et al. Evaluation of prostate cancer detection with ultrasound real-time elastography: a comparison with step section pathological analysis after radical prostatectomy. Euro Urol 2008; 54;6: 1354-62. Ultrasonic elastography in prostatic diseases-Guo et al. Urological Oncology 351