Publication: Evaluación de la fiabilidad de un programa de inteligencia artificial en el diagnóstico de caries sobre radiografías de aleta mordida.
Loading...
Official URL
Full text at PDC
Publication Date
2021
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citations
Exportar
Abstract
Introducción: En la actualidad, el diagnóstico precoz de la caries dental continúa siendo un desafío para los odontólogos. Aunque el método visual-táctil y la radiología se consideran métodos altamente fiables para la detección de la caries dental, el diagnóstico final tiende a verse influido por la experiencia subjetiva del operador. La inteligencia artificial puede hacer esas tareas de una manera sencilla, rápida y precisa, reduciendo el tiempo de trabajo y mejorando la toma de decisiones clínicas. Objetivo: Evaluar la fiabilidad de un programa de inteligencia artificial (DentiAI®), para la detección de caries interproximales en radiografías de aleta de mordida.
Material y Método: Para este trabajo se utilizaron 300 radiografías de aleta de mordida. La prueba de referencia establecida para este trabajo fue el diagnóstico de certeza establecido tras la realización de la exploración clínica, radiológica y la apertura de la cavidad correspondiente. Todas las radiografías válidas fueron sometidas a la evaluación del programa DentiAI®. Los resultados fueron analizados mediante la realización de un estudio estadístico basado en la medición de la potencia global del test, sensibilidad (S), especificidad (E), valor predictivo positivo (VPP), valor predictivo negativo (VPN), razón de probabilidad positiva LR (+) y razón de probabilidad negativa LR (-) de cuatro modelos del programa DentiAI® establecidos en función del valor del porcentaje que determina la presencia de caries, y su posterior representación en Curvas de Roc.
Resultados: Los modelos 1, 2, 3 y 4 mostraron los siguientes resultados espectivamente: Potencia global del test de 70.8%, 82%, 85.6%, 86.1%. S de 87%, 69.8%, 57%, 41.6%. E de 66.3%, 85.4%, 93.7%, 98.5%. VPP de 42%, 57.2%, 71.6%, 88.6%. VPN de 94.8%, 91%, 88.6%, 85.8%. LR(+) de 2.58, 4.78, 9.05, 27.73. LR(-) de 0.2, 0.35, 0.46, 0.59.Área bajo la curva ROC de 0.767, 0.777, 0.753, 0.701.
Conclusiones: El programa DentiAI® es una herramienta fiable en el diagnóstico radiológico de caries interproximales. El modelo 2 ofrece los mejores resultados. Los datos deben ser analizados en su conjunto para aprovechar las ventajas que nos ofrece cada modelo. La responsabilidad final es del profesional y nunca de la tecnología.
Introduction: Nowadays, the early diagnosis of dental caries continues to be a challenge for dentists. However, although the visual-tactile method and radiology are considered highly reliable methods for the detection of dental caries, the final diagnosis tends to be influenced by the subjective experience of the operator. Artificial intelligence can do these tasks in a simple, fast and precise way, reducing work time and improving clinical decision making. Objectives: To evaluate the reliability of an artificial intelligence program (DentiAI®), for the detection of interproximal caries in bitewing radiographs. Material and methods: For this work, 300 bitewing radiographs of patients were used. The reference test established for this work was the diagnosis established after carrying out the clinical and radiological examination, and the drilling of the cavity. All valid radiographs were subjected to the evaluation of the DentiAI® program. The results were stored and finally analyzed by conducting a statistical study based on the measurement of sensitivity (S), specificity (E), positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio LR (+) and negative likelihood ratio LR (-) of four models of the DentiAI® program, and its later representation in Roc Curves. Results: Models 1, 2, 3, 4 showed the following results respectively: Accuracy of 70.8%, 82%, 85.6%, 86.1%. S of 87%, 69.8%, 57%, 41.6%. E of 66.3%, 85.4%, 93.7%, 98.5%. PPV de 42%, 57.2%, 71.6%, 88.6%. NPV of 94.8%, 91%, 88.6%, 85.8%. LR(+) of 2.58, 4.78, 9.05, 27.73. LR(-) of 0.2, 0.35, 0.46, 0.59. Area under the ROC curve of 0.767, 0.777, 0.753, 0.701. Conclusions: The DentiAI® program is a reliable tool for the radiological diagnosis of interproximal caries. Model 2 offers the best results. The data must be analyzed as a whole. The final responsibility rests with the professional and never with the technology.
Introduction: Nowadays, the early diagnosis of dental caries continues to be a challenge for dentists. However, although the visual-tactile method and radiology are considered highly reliable methods for the detection of dental caries, the final diagnosis tends to be influenced by the subjective experience of the operator. Artificial intelligence can do these tasks in a simple, fast and precise way, reducing work time and improving clinical decision making. Objectives: To evaluate the reliability of an artificial intelligence program (DentiAI®), for the detection of interproximal caries in bitewing radiographs. Material and methods: For this work, 300 bitewing radiographs of patients were used. The reference test established for this work was the diagnosis established after carrying out the clinical and radiological examination, and the drilling of the cavity. All valid radiographs were subjected to the evaluation of the DentiAI® program. The results were stored and finally analyzed by conducting a statistical study based on the measurement of sensitivity (S), specificity (E), positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio LR (+) and negative likelihood ratio LR (-) of four models of the DentiAI® program, and its later representation in Roc Curves. Results: Models 1, 2, 3, 4 showed the following results respectively: Accuracy of 70.8%, 82%, 85.6%, 86.1%. S of 87%, 69.8%, 57%, 41.6%. E of 66.3%, 85.4%, 93.7%, 98.5%. PPV de 42%, 57.2%, 71.6%, 88.6%. NPV of 94.8%, 91%, 88.6%, 85.8%. LR(+) of 2.58, 4.78, 9.05, 27.73. LR(-) of 0.2, 0.35, 0.46, 0.59. Area under the ROC curve of 0.767, 0.777, 0.753, 0.701. Conclusions: The DentiAI® program is a reliable tool for the radiological diagnosis of interproximal caries. Model 2 offers the best results. The data must be analyzed as a whole. The final responsibility rests with the professional and never with the technology.
Description
UCM subjects
Unesco subjects
Keywords
Citation
Selwitz RH, Ismail AI, Pitts NB. Dental caries. Lancet. 2007;369(4):51–9.
2. Lee JH, Kim DH, Jeong SN, Choi SH. Detection and diagnosis of dental caries using a deep learning-based convolutional neural network algorithm.J Dent [Internet]. 2018;77(June):106–11. Available from: https://doi.org/10.1016/j.jdent.2018.07.015
3. Gomez J. Detection and diagnosis of the early caries lesion. BMC Oral Health. 2015;15(1):1–7.
4. MacHiulskiene V, Campus G, Carvalho JC, Dige I, Ekstrand KR, JablonskiMomeni A, et al. Terminology of Dental Caries and Dental Caries Management: Consensus Report of a Workshop Organized by ORCA and Cariology Research Group of IADR. Caries Res. 2020;54(1):7–14.
5. Pitts N. “ICDAS” - An international system for caries detection and assessment being developed to facilitate caries epidemiology, research and appropriate clinical management. Community Dent Health. 2004;21(3):193–8.
6. ICDAS Foundation: International Caries Detection and Assessment System. [Internet]. Available from: http://www.icdas.org.
7. Fusayama T. Two Layers of Carious Dentin: Diagnosis and Treatment. Oper Dent. 1946;4:63–70.
8. Chie S, Shigehisa I, Masafumi B, Hiroyasu H, Takao F. Caries detector for pulpless teeth. Oper Dent. 1983;8:94–8.
9. Tunc E Sen, Sahin Z, Guler E, Bayrak S, Tuloglu N. Effectiveness of Different Cleaning Solutions in Removing Residual Caries-Detector Dyes. J Esthet Restor Dent. 2019;31(3):253–6.
10. Unlu N, Ermis RB, Sener S, Kucukyilmaz E, Cetin AR. An In Vitro Comparison of Different Diagnostic Methods in Detection of Residual Dentinal Caries. Int J Dent. 2010;2010:1–8.
11. Tyas MJ, Anusavice KJ, Frencken JE, Mount GJ. Minimal intervention dentistry - A review: FDI Commission Project 1-97. Int Dent J. 2000;50(1):1–12.
12. Abogazalah N, Ando M. Alternative methods to visual and radiographic examinations for approximal caries detection. J Oral Sci. 2017;59(3):315–22.
13. Martínez-Rus F, García AM, de Aza AH, Pradíes G. Radiopacity of zirconia�based all-ceramic crown systems. Int J Prosthodont. 2011;24(2):144–6.
14. Espelid I, Tveit AB, Erickson RL, Keck SC, Glasspoole EA. Radiopacity of restorations and detection of secondary caries. Dent Mater. 1991;7(2):114–7.
15. Berry HM. Cervical burnout and Mach band: two shadows of doubt in radiologic interpretation of carious lesions. J Am Dent Assoc [Internet]. 1983;106(5):622–5. Available from:
http://dx.doi.org/10.14219/jada.archive.1983.0109
16. Ward NG. Automatic discovery of simply-composable prosodic elements. Proc Int Conf Speech Prosody. 2014;(4):915–9.
17. Ekstrand KR, Luna LE, Promisiero L, Cortes A, Cuevas S, Reyes JF, et al. The reliability and accuracy of two methods for proximal caries detection and depth on directly visible proximal surfaces: An in vitro study. Caries Res. 2011;45(2):93–9.
18. Ekstrand KR, Gimenez T, Ferreira FR, Mendes FM, Braga MM. The International Caries Detection and Assessment System - ICDAS: A Systematic Review. Caries Res. 2018;52(5):406–19.
19. Schwendicke F, Tzschoppe M, Paris S. Radiographic caries detection: A systematic review and meta-analysis. J Dent. 2015;43(8):924–33.
20. Prados-Privado M, García Villalón J, Martínez-Martínez CH, Ivorra C, Prados-Frutos JC. Dental Caries Diagnosis and Detection Using Neural Networks: A Systematic Review. J Clin Med. 2020;9(11):3579. 21. Foster Page LA, Boyd D, Fuge K, Stevenson A, Goad K, Sim D, et al. The effect of bitewing radiography on estimates of dental caries experience
among children differs according to their disease experience. BMC Oral Health. 2018;18(1):1–8.
22. Mejàre I, Källestål C, Stenlund H, Johansson H. Caries Development from 11 to 22 Years of Age: A Prospective Radiographic Study. Prevalence and Distribution. Caries Res. 1998;32(1):10–6.
23. Mejàre I, Källestål C, Stenlund H. Incidence and Progression of Approximal Caries from 11 to 22 Years of Age in Sweden: A Prospective Radiographic Study. Caries Res. 1999;33(2):93–100.
24. Tuzoff D V., Tuzova LN, Bornstein MM, Krasnov AS, Kharchenko MA, Nikolenko SI, et al. Tooth detection and numbering in panoramic radiographs using convolutional neural networks. Dentomaxillofacial
Radiol. 2019;48(4):1–11.
25. Review AQ. Computing machinery and intelligence. MIND. 1950;236:433–60.
26. Mintz Y, Brodie R. Introduction to artificial intelligence in medicine. Minim Invasive Ther Allied Technol [Internet]. 2019;28(2):73–81. Available from:
https://doi.org/10.1080/13645706.2019.1575882
27. Hamet P, Tremblay J. Artificial intelligence in medicine. Metabolism [Internet]. 2017;69:S36–40. Available from: http://dx.doi.org/10.1016/j.metabol.2017.01.011
28. Mupparapu M, Wu CW, Chen YC. Artificial intelligence, machine learning, neural networks, and deep learning: Futuristic concepts for new dental diagnosis. Quintessence Int (Berl). 2018;49(9):687–8.
29. Mazurowski MA, Buda M, Saha A, Bashir MR. Deep learning in radiology: An overview of the concepts and a survey of the state of the art with focus on MRI. J Magn Reson Imaging. 2019;49(4):939–54.
30. Schwendicke F, Golla T, Dreher M, Krois J. Convolutional neural networks for dental image diagnostics: A Scoping Review. J Dent [Internet].2019;104947. Available from: https://doi.org/10.1016/j.tiv.2020.104947
31. Esteva A, Kuprel B, Novoa RA, Ko J, Swetter SM, Blau HM, et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature. 2017;542(7639):115–8.
32. Gulshan V, Peng L, Coram M, Stumpe MC, Wu D, Narayanaswamy A, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA - J Am Med
Assoc. 2016;316(22):2402–10.
33. Geetha V, Aprameya KS, Hinduja DM. Dental caries diagnosis in digital radiographs using back-propagation neural network. Heal Inf Sci Syst. 2020;8(1).
34. Prajapati SA, Nagaraj R, Mitra S. Classification of dental diseases using CNN and transfer learning. 5th Int Symp Comput Bus Intell ISCBI 2017.
2017;70–4.
35. Becker AS, Marcon M, Ghafoor S, Wurnig MC, Frauenfelder T, Boss A. Deep learning in mammography diagnostic accuracy of a multipurpose image analysis software in the detection of breast cancer. Invest Radiol.
2017;52(7):434–40.
36. Garcia Cantu A, Sascha G, Krois J, Chaurasia A, Gómez Rossi J, Gaudin R, et al. Detecting caries lesions of different radiographic extension on bitewings using deep learning. J Dent [Internet]. 2020;102248. Available
from: https://doi.org/10.1016/j.janxdis.2020.102248
37. Grischke J, Johannsmeier L, Eich L, Griga L, Haddadin S. Dentronics: Towards robotics and artificial intelligence in dentistry. Dent Mater [Internet]. 2020;36(6):765–78. Available from: https://doi.org/10.1016/j.dental.2020.03.021
38. Chen YW, Stanley K, Att W. Artificial intelligence in dentistry: Current applications and future perspectives. Quintessence Int (Berl). 2020;51(3):248–57.
39. Thanathornwong B. Bayesian-based decision support system for assessing the needs for orthodontic treatment. Healthc Inform Res. 2018;24(1):22–8.
40. Patcas R, Bernini DAJ, Volokitin A, Agustsson E, Rothe R, Timofte R. Applying artificial intelligence to assess the impact of orthognathic treatment on facial attractiveness and estimated age. Int J Oral Maxillofac Surg [Internet]. 2019;48(1):77–83. Available from: https://doi.org/10.1016/j.ijom.2018.07.010
41. Lee JH, Kim DH, Jeong SN, Choi SH. Diagnosis and prediction of periodontally compromised teeth using a deep learning-based convolutional neural network algorithm. J Periodontal Implant Sci. 2018;48(2):114–23.
42. Feres M, Louzoun Y, Haber S, Faveri M, Figueiredo LC, Levin L. Support vector machine-based differentiation between aggressive and chronic periodontitis using microbial profiles. Int Dent J. 2018;68(1):39–46.
43. Srivastava MM, Kumar P, Pradhan L, Varadarajan S. Detection of tooth caries in bitewing radiographs using deep learning. arXiv. 2017;(Nips 2017).
44. Albahbah AA, El-Bakry HM, Abd-Elgahany S. Detection of Caries in Panoramic Dental X-ray Images using Back-Propagation Neural Network. Int J Electron Commun Comput Eng. 2016;7(5):250–6.
45. Singh P, Sehgal P. Automated caries detection based on Radon transformation and DCT. 8th Int Conf Comput Commun Netw Technol ICCCNT 2017. 2017;1–6.
46. Li W, Kuang W, Li Y, Li YJ, Ye WP. Clinical X-ray image based tooth decay diagnosis using SVM. Proc Sixth Int Conf Mach Learn Cybern ICMLC 2007. 2007;3(August):1616–9.
47. Yu Y, Li Y, Li YJ, Wang JM, Lin DH, Ye WP. Tooth decay diagnosis using back propagation neural network. Proc 2006 Int Conf Mach Learn Cybern.2006;2006(August):3956–9. 48. Silva Fuente-Alba C, Molina Villagra M. Likelihood ratio (razón de verosimilitud): definición y aplicación en Radiología. Rev Argentina Radiol. 2017;81(3):204–8.
49. Gansky SA. Dental Data Mining: Potential Pitfalls and Practical Issues S.A. 2003;17:109–14.
50. Downer M. Concurrent Validity of an Epidemiological Diagnostic System for Caries with the Histological Appearance of Extracted Teeth ad Validating Criterion. Caries Res [Internet]. 1975;53(9):231–46. Available
from: http://publications.lib.chalmers.se/records/fulltext/245180/245180.pdf%0A
https://hdl.handle.net/20.500.12380/245180%0Ahttp://dx.doi.org/10.1016/j
.jsames.2011.03.003%0Ahttps://doi.org/10.1016/j.gr.2017.08.001%0Ahttp
://dx.doi.org/10.1016/j.precamres.2014.12
51. Rigby AS. Statistical methods in epidemiology. V. Towards an understanding of the kappa coefficient. Disabil Rehabil. 2000;22(8):339–44.
52. McKinney SM, Sieniek M, Godbole V, Godwin J, Antropova N, Ashrafian H, et al. International evaluation of an AI system for breast cancer screening. Nature [Internet]. 2020;577(7788):89–94. Available from: http://dx.doi.org/10.1038/s41586-019-1799-6
53. Haenssle HA, Fink C, Schneiderbauer R, Toberer F, Buhl T, Blum A, et al. Man against Machine: Diagnostic performance of a deep learning convolutional neural network for dermoscopic melanoma recognition in
comparison to 58 dermatologists. Ann Oncol. 2018;29(8):1836–42