Original ArticleBraz J Oral Sci. April/June 2009 - Volume 8, Number 2 Digital versus conventional radiography for determination of primary incisor length Mariane Emi Sanabe1, Maria Daniela Basso2, Marcela Almeida Gonçalves3, Rita de Cássia Loiola Cordeiro4, Lourdes Santos-Pinto4 1DDS, MD, PhD, student in Pediatric Dentistry, Faculdade de Odontologia de Araraquara, Universidade Estadual Paulista “Júlio de Mesquita Filho” (Unesp), Araraquara (SP), Brazil 2DDS, MD, PhD, Associate professor, Department of Pediatric Dentistry, Cascavel Dental School, Universidade Estadual do Oeste do Paraná (Unioeste), Cascavel (PR), Brazil 3DDS, MD, PhD, Associate professor, Departament of Pediatric Dentistry, Universidade de Franca (UNIFRAN), Franca (SP), Brazil 4DDS, MD, PhD, Associate Professor, Department of Pediatric Dentistry, Faculdade de Odontologia de Araraquara, Unesp, Araraquara (SP), Brazil Received for publication: May 12, 2009 Accepted: July 24, 2009 Correspondence to: Lourdes Santos-Pinto Avenida Queiroz Filho, 972 – Vila Harmonia CEP 14802-610 – Araraquara (SP), Brazil E-mail: lspinto@foar.unesp.br Abstract Aim: This in vitro study evaluated the accuracy of primary incisor lengths determined by digital and conven- tional radiography compared to the actual tooth length. Methods: Twenty extracted primary maxillary incisors were mounted in acrylic blocks. Tooth length was estimated by using a straight-line measurement provided by the distance measurement tool of a digital dental imaging system (Computed Dental Radiography, Schick Technologies Inc.) and conventional E-speed film radiographs by using a digital caliper. Two operators familiar with both radiographic methods performed the estimates. The estimated tooth lengths were compared to the actual tooth lengths measured with the digital caliper. Data were statistically analyzed by Dahlberg’s equa- tion, paired t test, Pearson’s correlation coefficient and ANOVA at 5% significance level. Results: There were no statistically significant differences (p = 0.85) between tooth length estimated on digital and conventional radiographs. Admitting as clinically acceptable a 0.5-mm discrepancy between the actual tooth lengths and the radiographically estimated lengths, 60% of the radiographic measurements were considered as accurate. When the acceptable difference range was 1.0 mm, the accuracy of the radiographic measurements increased to 100%. Conclusions: Digital and conventional radiography provided similar tooth length measurements and were equivalent to the actual tooth lengths. Keywords: dental radiography, dental digital radiography, primary tooth, odontometry. Introduction Radiographic determination of tooth length is one of the critical aspects of pulpectomy in primary teeth because minor degrees of resorption may not be obvious radiographically1 and an underling permanent tooth germ can cause image superimposition. Consequently, the root apex is not clearly identified. In order to establish the correct working length (WL) for instrumentation of the root canal system, the tooth length should be estimated from a preoperative radiograph, an en- dodontic file should be inserted up to the established length and another radiograph should be taken to check whether the instrument is positioned at the right level. Therefore, accurate tooth length measurements are extremely important to ensure that the file does not pass be- yond the apical foramen and causes injuries to the periapical tissues2. Overinstrumentation in primary teeth involves the additional risk of damage to the permanent tooth germ3,4. 102 Sanabe ME, Basso MD, Gonçalves MA, Cordeiro RCL, Santos-Pinto LD Braz J Oral Sci. 8(2): 101-4 The WL is traditionally established based on the examination of conventional intraoral radiographs. The recent technological ad- vances turned the digital radiography into a viable option for the determination of the WL. The reliability of WLs established with use of digital radiography has been described as comparable to or even better than that of conventional radiography5-10. Other studies11-14, however, reported that conventional radiography is more accurate in comparison to older digital radiographic systems. For most digital radiographic systems, image acquisition and manipulation may be performed using the proprietary softwares supplied by the manufacturers or other commercially available graphic softwares. Among the options of digital radiographic sys- tems are the linear measurements between two or more spots local- ized on the image using the mouse, which produce a numerical mea- surement, generally in millimeters. Although this method has been used for WL determination in permanent teeth6,7,10, little is known about its reproducibility, consistence and accuracy for tooth length assessment in the primary dentition. The purpose of this study was to compare the accuracy of digital and conventional radiography as well as inter and intra-examiner agreement for determination of the length of primary incisors. Material and methods After approval of the research protocol by the Ethics Committee of Faculdade de Odontologia de Araraquara, children from the Pediat- ric Dentistry clinic with symptomatology and clinical findings of ne- crosis and radiographic diagnosis of periapical lesion in the primary incisors were invited to participate in this study. The informed con- sent was obtained from all parents/legal guardians. Twenty primary maxillary incisors with less than 2/3 of root resorption were used in this study. The teeth were stored in 1% thymol solution at room temperature until use. Each tooth was measured twice with a digital caliper (Mitu- toyo Corp., Tokyo, Japan) and the average was recorded as the ac- tual tooth length. The teeth were embedded in acrylic resin blocks and adapted to a custom-made radiograph positioning device that maintained the angle and distance (25 cm) constant between the radiation source and the tooth for both digital and conventional ra- diographs. A groove was made in the resin block to allow adaptation of the guiding shaft in order to adjust the cylinder of the X-ray equip- ment, thereby simulating the paralleling radiographic technique. The teeth were radiographed by a single calibrated operator. The exposure times were established in previous evaluations in which radiographs were obtained from extracted teeth with exposure set- ting ranging from 0.1 to 0.4 seconds. Two experienced radiologists, after a careful comparison of the quality of the images provided by both types of radiographic techniques, chose the most appropriates radiographic density for tooth length determination. The digital radiographs were obtained using a computed dental radiography system and #2 sensor (CDR-Schick Technologies Inc., Long Island City, NY, USA; version 2.6) and X-ray equipment (Gnatus XR 6010; Gnatus, Ribeirão Preto, SP, Brazil), operating at 60 kVp, 7 mA. The digital images were stored in TIFF format for further analy- sis. The conventional radiographs were taken with E-speed film (Eastman Kodak, Rochester, NY, USA) using the same X-ray equip- ment. The films were developed by the time/temperature method and mounted on slide frames. On the conventional radiographs, tooth length was estimated us- ing a digital caliper (Mitutoyo Corp.,). The radiographs were viewed on a light box with no magnification. On the digital radiographs, tooth length was measured directly on the screen of a high-resolu- tion 17” monitor with 100% zoom magnification. The measurement method was the electronic ruler of the proprietary CDR system soft- ware (version 2.6; Schick Technologies Inc.). Using the left mouse button, a two-click measurement was per- formed for tooth length determination: one click at the visible edge of the crown and the other at the root apex. Prior to the measurements, the electronic ruler was calibrated by measuring an object of known length, a #30 Kerr file (Les Fils D’Auguste Maillefer S.A., Switzerland). Enhancement features, such as brightness and contrast, were not used for the on-screen measurements. Two experienced pediatric dentists with expertise in both ra- diographic techniques were calibrated (kappa = 0.94) and assessed twice the length of the teeth with a two-month interval. Intra and inter-observer variability was determined by calcu- lating the error of the method from double measurements using Dahlberg’s equation and systematic errors were detected through a paired t test15. The measurements obtained with conventional radi- ography and digital radiography were compared to the actual tooth lengths and analyzed by ANOVA. A 5% significance level was set for all analyses. Results The analysis of intra-examiner agreement showed a high reproduc- ibility of tooth lengths for both examiners when digital and con- ventional radiographic measurements were performed (Table 1). Considering that both examiners presented excellent measurement reproducibility for both types of radiographic method and that there was small variation of tooth lengths and similar standard deviations, the means were used for results analysis. Inter-examiner agreement was excellent for digital radiographic images; however, measurements in the conventional radiography presented random and systematic error (Table 1). There was no statistically significant difference (p = 0.85) be- tween the actual tooth lengths and the tooth lengths for both radio- graphic image types. However, overestimation of the actual tooth length occurred in 70% of the digital radiographs and in 75% of the conventional radiographs (Table 2). Admitting a 0.5-mm difference between the actual tooth lengths and lengths estimated on the radiographs as a clinically acceptable 103Digital versus conventional radiography for determination of primary incisor length Braz J Oral Sci. 8(2): 101-4 Intra-examiners Inter-examiners Rx D Rx C Rx D Rx C Ex 1 Ex 2 Ex 1 Ex 2 Mean (sd) 11.16 (1.46) 11.26 (1.54) 11.06 (1.55) 11.36 (1.52) 11.21 (1.48) 11.21 (1.53) Dalberg’s 0.06 0.04 0.00 0.01 -0.10 -0.29 p 0,469 0,494 0,959 0,695 0,319 .001* r 0.95 0.97 0.97 0.98 0.96 0.97 Table 1. Tooth length means and standard deviations (mm), random error (Dalberg’s equation), systematic error (p) and Pearson’s correlation coefficient (r) for intra and inter-examiners (Ex) reliability for digital (Rx D) and conventional (Rx C) images sd: standard deviation; *Diferença estatística. Radiography Overestimated (%) Underestimated (%) Maximun (mm) Minimun (mm) Mean (mm) Digital 70 30 +0.72 - 0.90 0.40 Conventional 75 25 +0.85 -0.90 0.44 Table 2. Percentage of overestimated and underestimated radiographic tooth lengths as well as maximum, minimum and mean discrepancies from the actual tooth lengths for each radiographic method discrepancy, 60% of the measurements were considered accurate for both digital and conventional radiography. When the acceptable difference range was 1.0 mm, the accuracy of the radiographic mea- surements increased to 100%. Discussion The results of this study showed no significant difference between the primary tooth lengths obtained with digital and conventional radiographs, suggesting that digital radiography was effective in the assessment of primary anterior tooth length. These findings are compliant with those of similar studies that compared the accuracy of digital images and conventional radiographs for determination of the length primary teeth in vivo16, extracted primary teeth in vitro17 and extracted permanent teeth6,7,10 . The reproducibility of the measurements after a 2-month inter val was excellent for both examiners using either ty pes of radiographic image (Table 1). An excellent correlation between digital and conventional radiography has also been found for estimation of canal leng th in permanent teeth w ith a two-week inter val between image acquisition18 as well as for assessment of endodontic W Ls in permanent molars obtained from human ca- daver ma xilla19,20. The f indings of the present study revealed excellent agree- ment between the examiners for digital radiographic images measured directly on the screen monitor w ith 100% zoom mag- nif ication. However, the performance of the examiners for con- ventional image presented random error (Dalberg’s error = 0.29) and systematic error (p = 0.01), probably inf luenced by the radio- graphs v iew w ith no magnif ication and diff iculty to identif y the root apex when the root resorption was irreg ular. The results of the present study do not agree w ith those of Ellingsen et al.11, who found high inter-examiner agreement in both radiov isiography and conventional radiography for determining the position of the tips of small-sized f iles relative to the radiographic apex in ex- tracted permanent teeth. Accepting clinically a 0.5-mm discrepancy between the actual tooth length and lengths estimated on the radiographs, 60% of the measurements obtained with either types of radiographic image were considered equivalent to the actual tooth lengths. These findings are consistent with those of Martinez-Martinez-Lozano et al.6, who radio- graphically compared estimate and actual tooth lengths admitting as acceptable a 0.5-mm difference range, and found that conventional and digital radiological methods were satisfactory in 50.6 and 61.4% of cases, respectively. Leddy et al.19 compared RadioVisioGraphy imaging and conventional film-based radiography in determining endodontic file length adjustment and reported a 50% increase in length adjust- ment estimates when a 0.5-mm difference between radiographic and actual measurements were considered as satisfactory. In the present study, when the acceptable difference range was 1.0 mm, the accuracy of the radiographic measurements increased to 100%. This result corroborates the findings of Larheim and Eg- gen21, who observed that, for conventional radiography, 95% of the sample presented WL estimates equivalent to the actual WLs, ad- mitting a 1.0-mm variation. Most tooth length measurements were overestimated in less than 1 mm for both digital and conventional film-based radiographs. This discrepancy could be considered as clinically acceptable because, in order to determine the WL, a radiograph is taken with an endodontic file placed inside the canal 2 mm short of the tooth length. The 1 mm overestimated measurements observed in this study did not allow the file to extend beyond the actual tooth length and past the api- cal foramen. Slightly overestimated root lengths have also been ob- served in digital radiographs of extracted permanent teeth7,18,22 and dry mandibular jaw sections23. This study controlled possible sources of error in radiographic images, such as the distance from the tooth to the radiation source and to the film or sensor, as well as the vertical and horizontal cone angulation. A custom-made positioning device allowed holding the specimen, the X-ray equipment cone and the film and maintaining the sensor in a fixed position. It was, therefore, possible to simulate the paralleling radiographic technique, in which the estimated tooth length is closer to the actual length21. 104 Sanabe ME, Basso MD, Gonçalves MA, Cordeiro RCL, Santos-Pinto LD Braz J Oral Sci. 8(2): 101-4 Digital image calibration was performed before each tooth length determination using the on-screen calibration tool to meas- ure the image of an endodontic file of a known length. It was done because it has been shown that calibrated digital measurements are more accurate than uncalibrated measurements24. In conclusion, digital and conventional radiography prov ided similar tooth leng th measurements and were equivalent to the actual tooth leng ths, which validates both techniques for endo- dontic image acquisition in primar y teeth. The null hy pothesis is therefore accepted. It must be emphasized, however, that in this study the radiographic images were obtained under well controlled and standardized conditions, which is ver y diff icult to be reproduced in a clinical setting, especially w ith pediatric patients. Notw ithstanding, some characteristics of the digital im- ages, such as reduced patient exposure to radiation, possibility of producing instant images and elimination of the chemical solu- tions used for image processing, make them a valuable alternative for endodontic leng th measurements in primar y teeth. Acknowledgements The authors wish to thank to Fundação de Amparo à Pesquisa do Estado de São Paulo (Fapesp) for the financial support. References 1. Mente J, Seidel J, Buchalla W, Koch MJ. Electronic determination of root canal length in primary teeth with and without root resorption. Int Endod J. 2002;35:447-52. 2. Rodd HD, Waterhouse PJ, Fuks AB, Fayle SA, Moffar MA. Pulp therapy for primary molars. Int J Paed Dent. 2006;16:15-23. 3. Troutman KC, Reisbick MH. Pulp therapy. In: Stewart RE, Barber TK, Troutman KC, Wei SHY, editors. Pediatric Dentistry: Scientific foundations and clinical practice. 1982; Toronto: Mo Mosby Co. p. 908-17. 4. Katz A, Mass E, Kaufman AY. Electronic apex locator: a useful tool for root canal treatment in the primary dentition. ASDC J Dent Child. 1996;63:414-7. 5. Cedeberg RA, Tidwell E, Frederiksen NL, Benson BW. Endodontic working length assessment: Comparison of storage phosphor digital imaging and radiographic film. Oral Surg Oral Med Oral Pathol Endod. 1998;85:325-8. 6. Martinez-Lozano MA, Forner-Navarro L, Sanches-Cortez JL, Llena-Puy C. Methodological considerations in the determination of working length. Int Endod J. 2001;34:371-6. 7. Mentes A, Gencoglu N. Canal length evaluation of curved canals by direct digital or conventional radiography. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2002;93:88-91. 8. Radel RT, Goodell GG, McClanaban SB, Cohen ME. In vitro radiographic determination of distances from working length files to tooth ends comparing Kodak RVG 6000, Schick, and Kodak Insight film. J Endod. 2006;32:566-8. 9. Versteeg KH, Sanderink GCH, van Ginkel FC, van der Stelt PF. Estimating distances on direct digital images and conventional radiographs. J Am Dent Assoc. 1997;128:439-43. 10. Woolhiser GA, Brand JW, Hoen MM, Geist JR, Pikula AA, Pink FE. Accuracy of film- based, digital, and enhanced digital images for endodontic length determination. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2005;99:499-504. 11. Ellingsen MA, Harrington GW, Hollender LG. RadioVisioGraphy versus conventional radiography for detection of small instruments in endodontic length determination part I: In vitro evaluation. J Endod. 1995;21:326-31. 12. Ellingsen MA, Hollender LG, Harrington GW. RadioVisioGraphy versus conventional radiography for detection of small instruments in endodontic length determination part II: In vivo evaluation. J Endod. 1995;21:516-20. 13. Hedrick RT, Dove SB, Peters DD, McDavid WD. Radiographic determination of canal length: direct digital radiography versus conventional radiography. J Endod. 1994;20:320-6. 14. Sanderink GCH, Huiskens R, van der Stelt PF, Welander US, Stheeman SE. Image quality of direct digital intraoral x-ray sensors in assessing root canal length. The RadioVisioGraphy, Visualix/VIXA, Sens-A-Ray, and Flash Dent systems compared with Ektaspeed films. Oral Surg Oral Med Oral Pathol. 1994;78:125-32. 15. Houston, WJB. The analysis of errors in orthodontic measurements. Am J Orthod.1983;83:383-90. 16. Santos-Pinto L, Cordeiro R de C, Zuanon AC, Basso MD, Gonçalves MA. Primary tooth length determination in direct digital radiography: an in vivo study. Pediatr Dent. 2007;29:470-4. 17. Subramaniam P, Konde S, Mandanna DK. An in vitro comparison of root canal measurement in primary teeth. J Indian Soc Pedod Prev Dent. 2005;23:124-5. 18. Burger CL, Mork TO, Hutter JW, Nicoll B. Direct digital radiography versus conventional radiography for estimation of canal length in curved canals. J Endod. 1999;25:260-3. 19. Leddy BJ, Miles DA, Newton CW, Brown CE. Interpretation of endodontic file lengths using RadioVisioGraphy. J Endod. 1994;20:542-5. 20. Sheaffer JC, Eleazer PD, Schetz JP, Clark SJ, Farman AG. Endodontic measurement accuracy and perceived radiograph quality: Effects of film speed and density. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2003;96:441-8. 21. Larheim TA, Eggen S. Determination of tooth length with a standardized paralleling technique and calibrated radiographic measuring film. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1979;48:374-8. 22. Velders XL, Sanderink GCH, van der Stelt PF. Dose reduction of two digital sensor systems measuring file lengths. Oral Surg Oral Med Oral Pathol Endod. 1996;81:607-12. 23. Borg E, Gröndahl HG. Endodontic measurement in digital radiographs acquired by a photostimulable, storage phosphor system. Endod Dent Traumatol. 1996;12:20-4. 24. Loushine RJ, Weller N, Kimbrough WF, Potter BJ. Measurement of endodontic file lengths: calibrated versus uncalibrated digital images. J Endod. 2001;27:779-81.