Title

In vitro assessment of AI-driven prediction for cyclic fatigue failure in NiTi rotary files

Authors

Swagat Panda*, Chinmayee Priyadarsini, Ayesha Satapathy, Eleena Satapathy, Jasasriya Nanda & Sushree Soumya Suravi

Affiliation

Department of Conservative Dentistry and Endodontics, Hi –Tech Dental College & Hospital ,Bhubaneswar -751025, Odisha, India; *Corresponding author

Email

Swagat Panda - E-mail: reachswagat@gmail.com
Chinmayee Priyadarsini - E-mail: chinmayeepriyadarsini46@gmail.com

Ayesha Satapathy - E-mail: drayeshasatapathy@gmail.com
Eleena Satapathy - E-mail: satapathyeleena@gmail.com
Jasasriya Nanda - E-mail: jasasriyananda@gmail.com
Sushree Soumya Suravi - E-mail: sushree.suravi123@gmail.com

Article Type

Research Article

Date

Received November 15, 2025; Revised December 15, 2025; Accepted December 15, 2025, Published December 15, 2025

Abstract

Nickel-titanium (NiTi) rotary files are also a significant clinical problem, which results in the separation of instruments during endodontic work due to cyclic fatigue failure. Therefore, it is of interest validate an AI predictive model by utilizing real-time operational parameters of five NiTi file systems. A CNN-LSTM was used to predict future failure based on the data on torque, angular velocity, vibration, and temperature. The model had a high level of accuracy at 94.2 and sensitivity and specificity, which was better than the methods of traditional prediction. Monitoring using AI is a promising solution to prevent NiTi files fracture in real-time to make the endodontic process safer.

Keywords

Artificial intelligence, machine learning, nickel-titanium, rotary files, cyclic fatigue, endodontics, predictive modeling, deep learning

Citation

Panda et al. Bioinformation 21(12): 4825-4830 (2025)

Edited by

Hiroj Bagde

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.