Jennifer Chang1, Ming-Feng Chang2,3, Nikola Angelov4, Chih-Yu Hsu2, Hsiu-Wan Meng4, Sally Sheng4, Aaron Glick5, Kearny Chang4, Yun-Ru He2, Yi-Bing Lin2,3, Bing-Yan Wang4, Srinivas Ayilavarapu4. 1. Department of Periodontics and Dental Hygiene, The University of Texas Health Science Center at Houston School of Dentistry, Houston, TX, USA. Jennifer.chang@uth.tmc.edu. 2. Institute of Computational Intelligence, National Yangming Chiaotung University, Taipei, Taiwan. 3. Department of Computer Science, National Yangming Chiaotung University, Taipei, Taiwan. 4. Department of Periodontics and Dental Hygiene, The University of Texas Health Science Center at Houston School of Dentistry, Houston, TX, USA. 5. Department of General Practice and Dental Public Health, The University of Texas Health Science Center at Houston School of Dentistry, Houston, TX, USA.
Abstract
OBJECTIVE: Successful application of deep machine learning could reduce time-consuming and labor-intensive clinical work of calculating the amount of radiographic bone loss (RBL) in diagnosing and treatment planning for periodontitis. This study aimed to test the accuracy of RBL classification by machine learning. MATERIALS AND METHODS: A total of 236 patients with standardized full mouth radiographs were included. Each tooth from the periapical films was evaluated by three calibrated periodontists for categorization of RBL and radiographic defect morphology. Each image was pre-processed and augmented to ensure proper data balancing without data pollution, then a novel multitasking InceptionV3 model was applied. RESULTS: The model demonstrated an average accuracy of 0.87 ± 0.01 in the categorization of mild (< 15%) or severe (≥ 15%) bone loss with fivefold cross-validation. Sensitivity, specificity, positive predictive, and negative predictive values of the model were 0.86 ± 0.03, 0.88 ± 0.03, 0.88 ± 0.03, and 0.86 ± 0.02, respectively. CONCLUSIONS: Application of deep machine learning for the detection of alveolar bone loss yielded promising results in this study. Additional data would be beneficial to enhance model construction and enable better machine learning performance for clinical implementation. CLINICAL RELEVANCE: Higher accuracy of radiographic bone loss classification by machine learning can be achieved with more clinical data and proper model construction for valuable clinical application.
OBJECTIVE: Successful application of deep machine learning could reduce time-consuming and labor-intensive clinical work of calculating the amount of radiographic bone loss (RBL) in diagnosing and treatment planning for periodontitis. This study aimed to test the accuracy of RBL classification by machine learning. MATERIALS AND METHODS: A total of 236 patients with standardized full mouth radiographs were included. Each tooth from the periapical films was evaluated by three calibrated periodontists for categorization of RBL and radiographic defect morphology. Each image was pre-processed and augmented to ensure proper data balancing without data pollution, then a novel multitasking InceptionV3 model was applied. RESULTS: The model demonstrated an average accuracy of 0.87 ± 0.01 in the categorization of mild (< 15%) or severe (≥ 15%) bone loss with fivefold cross-validation. Sensitivity, specificity, positive predictive, and negative predictive values of the model were 0.86 ± 0.03, 0.88 ± 0.03, 0.88 ± 0.03, and 0.86 ± 0.02, respectively. CONCLUSIONS: Application of deep machine learning for the detection of alveolar bone loss yielded promising results in this study. Additional data would be beneficial to enhance model construction and enable better machine learning performance for clinical implementation. CLINICAL RELEVANCE: Higher accuracy of radiographic bone loss classification by machine learning can be achieved with more clinical data and proper model construction for valuable clinical application.
Authors: Paul I Eke; Bruce A Dye; Liang Wei; Gary D Slade; Gina O Thornton-Evans; Wenche S Borgnakke; George W Taylor; Roy C Page; James D Beck; Robert J Genco Journal: J Periodontol Date: 2015-02-17 Impact factor: 6.993
Authors: Iain L C Chapple; Brian L Mealey; Thomas E Van Dyke; P Mark Bartold; Henrik Dommisch; Peter Eickholz; Maria L Geisinger; Robert J Genco; Michael Glogauer; Moshe Goldstein; Terrence J Griffin; Palle Holmstrup; Georgia K Johnson; Yvonne Kapila; Niklaus P Lang; Joerg Meyle; Shinya Murakami; Jacqueline Plemons; Giuseppe A Romito; Lior Shapira; Dimitris N Tatakis; Wim Teughels; Leonardo Trombelli; Clemens Walter; Gernot Wimmer; Pinelopi Xenoudi; Hiromasa Yoshie Journal: J Periodontol Date: 2018-06 Impact factor: 6.993
Authors: J M Noble; L N Borrell; P N Papapanou; M S V Elkind; N Scarmeas; C B Wright Journal: J Neurol Neurosurg Psychiatry Date: 2009-05-05 Impact factor: 10.154