aceTrajectory-aware risk stratification of oral lichen planus using a multimodal large language model: a longitudinal diagnostic accuracy study
To evaluate the performance of a multimodal large language model (LLM) for longitudinal trajectory classification and risk stratification of oral lichen planus (OLP), compared with expert panel consensus. This retrospective diagnostic accuracy study included 300 patients with histopathologically confirmed OLP and at least 24 months of follow-up. Multimodal longitudinal case profiles (serial clinical records, intraoral photographs, and histopathology reports) were independently assessed by (ChatGPT, OpenAI) and a…
In 300 patients with histopathologically confirmed OLP and at least 24 months follow-up, a multimodal LLM achieved 94.7% trajectory classification accuracy and 99.6% specificity for detecting expert-defined high-risk cases.
The model missed 21.2% of expert-defined high-risk cases with sensitivity of 78.8%, and three-level risk stratification accuracy was only 76.3% with 98.6% of errors being downward shifts that underestimate risk.
Retrospective single-cohort design without prospective external validation; findings limited to histopathologically confirmed OLP cases with at least 24 months follow-up and may not generalize.
Evidence
- Peer-reviewedScientific Reports2026-08-13
How should this claim be treated?
Truvace Impact Record TRV-2026-0788, v1: “Trajectory-aware risk stratification of oral lichen planus using a multimodal large language model: a longitudinal diagnostic accuracy study.” Truvace, 2026-08-16. /record/TRV-2026-0788 (accessed at citation time). sha256 148cf95084751790…
Calibration history
Every change to this record since certification, in the open. None yet — the reading has held since it entered the record.
Certified into the record
How to verify without trusting this page
Fetch the canonical text of any version from /api/record/TRV-2026-0788 and hash it yourself — for example shasum -a 256 on the saved canonical field. The result must equal content_hash, and each version’s text ends with prev:followed by the prior version’s hash (version 1 chains to 64 zeros). If a single character of any version had been altered since certification, the chain would not reproduce.