What question did this study set out to answer?

To evaluate multimodal vision-language AI models for triaging endodontic treatment while tolerating image errors.

February 2, 2026International Journal of Dentistry0 citationsOpen Access

Error‐Tolerant Multimodal Vision‐Language Models for Endodontic Triaging: A Cross‐Sectional Study

MTMd Fahim Shahoriar TituNorth South University MPMahir Afser PavelNorth South University AAAfifa Zain ApurbaNorth South University

Key Points

To evaluate multimodal vision-language AI models for triaging endodontic treatment while tolerating image errors.
Collected 3600 dental images over one year
Applied image augmentation techniques for model enhancement
Fine-tuned models including BLIP, CLIP, Florence 2, and Paligemma using quantisation-aware training
Evaluated performance with BLEU, METEOR, ROUGE, CIDEr, and convergence metrics.
Quantisation-aware optimisation improved BLEU-4 by at least 17.3% and METEOR by 11.1%
ROUGE-L improved by 9.8% and CIDEr increased by 75.5% across all models
Memory consumption reduced by 87.5%, with 90% accuracy in triage assessments within 0.5% error margin.

Abstract

Objective To assess the performance of multimodal vision‐language artificial intelligence models, optimised using quantisation‐aware training, in triaging endodontic treatment needs. The focus is on the ability to interpret endodontic radiographs while tolerating common image capture errors, including cone cutting, elongation, foreshortening, horizontal misalignment, over‐ and under‐exposure and artefacts. Methods In total 3600 dental were obtained across a 1‐year period. Image augmentation techniques were applied to enhance model generalisability. Bootstrapped Language‐Image Pretraining (BLIP), CLIP, Florence 2 and Paligemma multimodal models were fine‐tuned using quantisation‐aware training and evaluated using Bilingual Evaluation Understudy (BLEU), Recall‐Oriented Understudy for Gisting Evaluation (ROUGE), Metric for Evaluation of Translation with Explicit ORdering (METEOR), Consensus‐based Image Description Evaluation (CIDEr) and Loss Trends and Convergence. Results Quantisation‐aware optimisation improved BLEU‐4 by at least 17.3%, METEOR by at least 11.1%, ROUGE‐L by at least 9.8% and CIDEr by at least 75.5% across all models. Quantisation reduced memory consumption by 87.5% while preserving diagnostic accuracy within a 0.5% error margin while correctly reproducing over 90% endodontic triage assessments made by practitioners. Conclusion Multimodal AI demonstrates tolerance to imaging inconsistencies and is capable of accurately triaging endodontic cases with minimal computational demands, without compromising diagnostic performance.

Ask AI

Helpful

Bookmark

View Full Paper