HOME   |    PDF   |   


Title

Artificial intelligence versus oral radiologists for TMJ osteoarthritis detection on CBCT

 

Authors

Haider Iqbal 1,*, Ankit Saha 2, Shivam Yadav 3, Syed Sajid Basha 4, Bassam Alkhalifah 5 & Love Kumar Bhatia 6

 

Affiliation

1Department of Oral Medicine and Radiology, Sardar Patel Post Graduate Institute of Dental and Medical Sciences, Lucknow , Uttar Pradesh, India; 2Department of Oral Medicine and Radiology, Kantipur Dental College Teaching Hospital and Research Centre, Kathmandu, Nepal, South Asia; 3Department of Dentistry, Autonomous State Medical College, Hardoi, Uttar Pradesh, India; 4Department of Oral and Maxillofacial Surgery, College of Dental Science and Hospital, Amargadh, Bhavnagar, Gujarat, India; 5Department of Radiology, College of Medicine, Qassim University, Buraydah, Saudi Arabia; 6Department of Prosthodontics, Sardar Patel Post Graduate Institute of Dental and Medical Sciences, Lucknow , Uttar Pradesh, India; *Corresponding author

 

Email

Haider Iqbal - E-mail: haideriqb@gmail.com
Ankit Saha - E-mail: ankit.omdr@gmail.com
Shivam Yadav - E-mail: shivam0710@gmail.com
Syed Sajid Basha - E-mail: sajidaahil@gmail.com
Bassam Alkhalifah - E-mail: b.alkhalifah@qu.edu.sa
Love Kumar Bhatia - E-mail: lovebhatia87@gmail.com
 

Article Type

Research Article

 

Date

Received October 1, 2026; Revised October 31, 2026; Accepted October 31, 2026, Published October 31, 2026

 

Abstract

Temporomandibular joint osteoarthritis (TMJ-OA) presents heterogeneous CBCT features that may be challenging to interpret consistently across readers. Therefore, it is of interest to evaluate a YOLO-based deep-learning model against oral radiologists for detecting and classifying TMJ osteoarthritic changes in 420 joints. On the locked test set, the AI model achieved an AUC of 0.93, sensitivity of 88.5% and specificity of 86.8% for overall osteoarthritis detection. Its performance was comparable to experienced radiologists and exceeded that of the junior reader for selected features, while AI assistance improved junior-reader sensitivity from 76.9% to 86.5%. Thus, data shows the use of AI as an effective second-reader tool to enhance CBCT interpretation rather than as an autonomous diagnostic replacement.

 

Keywords

Temporomandibular joint (TMJ), osteoarthritis, CBCT, artificial intelligence, deep learning, diagnostic accuracy

 

Citation

Iqbal et al. Bioinformation 22(10): 7522-7526 (2026)

 

Edited by

Ashwini Dhopte

 

ISSN

0973-2063

 

Publisher

Biomedical Informatics

 

License

This is an Open Access article which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. This is distributed under the terms of the Creative Commons Attribution License.