PulseTrendingJournal ClubResearchersJournalsExplore
Instagram
HomeTrendingJournal ClubExplore
Synapse
⌘+K
Synapse
March 21, 2026Cell Reports MedicineOpen Access

Developing and evaluating multimodal large language model for orthopantomography analysis to support clinical dentistry

View Full Paper
Ask AI
Bookmark
Share

Authors

XLXinyu LiuKHKuo Feng HungWYWeihao Yu

Discussion

Loading...

Member takes

Overview

Evaluating a diagnostic system for dental analysis, highlighting improved interactivity and performance.

Key Points

  • The aim is to develop and assess ToothXpert, a multimodal large language model for dental diagnostics using orthopantomography.
  • Developed a dental concept alignment dataset and MM-OPG dataset with 131,065 question-answer pairs.
  • Implemented ToothXpert to provide visual and language responses for OPG analysis.
  • Evaluated performance using internal and external test datasets.
  • Achieved a macro F1 score of 78.61% on the internal test dataset, outperforming several existing models.
  • Human assessment rated ToothXpert's performance at 3.54, significantly higher than other large language models.
  • On the external dataset, ToothXpert surpassed junior dentists' F1 scores by 1.96% and 2.99%, while requiring less diagnostic time.

Cite This Study

Liu et al. (2026) studied this question.

synapsesocial.com/papers/69be34d16e48c4981c672f3bhttps://doi.org/10.1016/j.xcrm.2026.102652
View Full Paper
Ask AI
Bookmark
Share