Fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographs

Introduction: Mandibular third molars (MTMs) are the most frequently impacted teeth, making their detection and classification essential before surgical extraction. This study aims to develop and assess the accuracy of a deep learning model for detecting and classifying impacted mandibular third mol...

Full description

Saved in:
Bibliographic Details
Main Authors: Veerabhadrappa Suresh Kandagal, Vengusamy Sivakumar, Padarha Shreyansh, Iyer Kiran, Yadav Seema
Format: Article
Language:English
Published: EDP Sciences 2025-01-01
Series:Journal of Oral Medicine and Oral Surgery
Subjects:
Online Access:https://www.jomos.org/articles/mbcb/full_html/2025/01/mbcb240248/mbcb240248.html
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849735981623672832
author Veerabhadrappa Suresh Kandagal
Vengusamy Sivakumar
Padarha Shreyansh
Iyer Kiran
Yadav Seema
author_facet Veerabhadrappa Suresh Kandagal
Vengusamy Sivakumar
Padarha Shreyansh
Iyer Kiran
Yadav Seema
author_sort Veerabhadrappa Suresh Kandagal
collection DOAJ
description Introduction: Mandibular third molars (MTMs) are the most frequently impacted teeth, making their detection and classification essential before surgical extraction. This study aims to develop and assess the accuracy of a deep learning model for detecting and classifying impacted mandibular third molars (IMTMs) using panoramic radiographs (PRs). Materials and methods: The study utilized a dataset of 1100 PRs with 1200 IMTMs and 711 PRs without MTMs. An oral radiologist validated the annotations, and the data were split into training, validation, and testing sets. The Sobel Third Molar Detection Model (STMD), built on the VGG16 architecture, identified MTMs. Detected MTMs were located using the YOLOv7 model and classified per Winter’s classification via a ResNet50-based prediction model. Results: The VGG16-based detection model achieved a testing accuracy of 93.51%, with a precision of 94.64, recall of 89.47, and an F1 score of 91.97. The ResNet50-based classification model attained a testing accuracy of 92.17%, precision of 92.1, recall of 92.17, and an AUC of 98.28. These findings demonstrate the high accuracy and reliability of both models. Conclusion: VGG16 and ResNet50 integrated with YOLOv7, demonstrated high accuracy suggesting that the automatic detection and classification of IMTMs can be significantly improved using these models.
format Article
id doaj-art-9c38f7e10461412ca43a9fd0765217b6
institution DOAJ
issn 2608-1326
language English
publishDate 2025-01-01
publisher EDP Sciences
record_format Article
series Journal of Oral Medicine and Oral Surgery
spelling doaj-art-9c38f7e10461412ca43a9fd0765217b62025-08-20T03:07:24ZengEDP SciencesJournal of Oral Medicine and Oral Surgery2608-13262025-01-01311710.1051/mbcb/2025008mbcb240248Fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographsVeerabhadrappa Suresh Kandagal0https://orcid.org/0000-0003-3784-594XVengusamy Sivakumar1Padarha Shreyansh2Iyer Kiran3https://orcid.org/0000-0003-3551-870XYadav Seema4https://orcid.org/0000-0002-9773-6641Department of Oral Diagnostic Sciences, Faculty of Dentistry, SEGi UniversitySchool of Computing, Faculty of Computing and Engineering Technology, Asia Pacific University of Technology and Innovation (APU)Oxford Internet Institute, University of OxfordPreventive Dental Science Department, College of Dentistry, King Saud Bin Abdulaziz University for Health SciencesDepartment of Periodontics and Implantology, Faculty of Dentistry, SEGi UniversityIntroduction: Mandibular third molars (MTMs) are the most frequently impacted teeth, making their detection and classification essential before surgical extraction. This study aims to develop and assess the accuracy of a deep learning model for detecting and classifying impacted mandibular third molars (IMTMs) using panoramic radiographs (PRs). Materials and methods: The study utilized a dataset of 1100 PRs with 1200 IMTMs and 711 PRs without MTMs. An oral radiologist validated the annotations, and the data were split into training, validation, and testing sets. The Sobel Third Molar Detection Model (STMD), built on the VGG16 architecture, identified MTMs. Detected MTMs were located using the YOLOv7 model and classified per Winter’s classification via a ResNet50-based prediction model. Results: The VGG16-based detection model achieved a testing accuracy of 93.51%, with a precision of 94.64, recall of 89.47, and an F1 score of 91.97. The ResNet50-based classification model attained a testing accuracy of 92.17%, precision of 92.1, recall of 92.17, and an AUC of 98.28. These findings demonstrate the high accuracy and reliability of both models. Conclusion: VGG16 and ResNet50 integrated with YOLOv7, demonstrated high accuracy suggesting that the automatic detection and classification of IMTMs can be significantly improved using these models.https://www.jomos.org/articles/mbcb/full_html/2025/01/mbcb240248/mbcb240248.htmlclassificationdeep learningimpacted teethmandibleradiographsthird molar
spellingShingle Veerabhadrappa Suresh Kandagal
Vengusamy Sivakumar
Padarha Shreyansh
Iyer Kiran
Yadav Seema
Fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographs
Journal of Oral Medicine and Oral Surgery
classification
deep learning
impacted teeth
mandible
radiographs
third molar
title Fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographs
title_full Fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographs
title_fullStr Fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographs
title_full_unstemmed Fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographs
title_short Fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographs
title_sort fully automated deep learning framework for detection and classification of impacted mandibular third molars in panoramic radiographs
topic classification
deep learning
impacted teeth
mandible
radiographs
third molar
url https://www.jomos.org/articles/mbcb/full_html/2025/01/mbcb240248/mbcb240248.html
work_keys_str_mv AT veerabhadrappasureshkandagal fullyautomateddeeplearningframeworkfordetectionandclassificationofimpactedmandibularthirdmolarsinpanoramicradiographs
AT vengusamysivakumar fullyautomateddeeplearningframeworkfordetectionandclassificationofimpactedmandibularthirdmolarsinpanoramicradiographs
AT padarhashreyansh fullyautomateddeeplearningframeworkfordetectionandclassificationofimpactedmandibularthirdmolarsinpanoramicradiographs
AT iyerkiran fullyautomateddeeplearningframeworkfordetectionandclassificationofimpactedmandibularthirdmolarsinpanoramicradiographs
AT yadavseema fullyautomateddeeplearningframeworkfordetectionandclassificationofimpactedmandibularthirdmolarsinpanoramicradiographs