Truace
Subscribe

AI read mouth lesion photos with 97.1% accuracy but got urgency right in just 70%

Study title: Image-Based Diagnosis of Oral Lesions: Performance of a Vision-Language Model versus Human Clinicians

Objective To evaluate the real-world diagnostic performance of a multimodal large language model (LLM) for image-based assessment of oral mucosal lesions compared with clinicians of varying expertise. Study design Prospective international multicenter diagnostic accuracy study.

Read the study abstract

Objective To evaluate the real-world diagnostic performance of a multimodal large language model (LLM) for image-based assessment of oral mucosal lesions compared with clinicians of varying expertise. Study design Prospective international multicenter diagnostic accuracy study. Setting Twenty university and tertiary head and neck centers in Italy, Belgium, France, Spain, and Israel. Methods We enrolled 350 consecutive patients (320 with oral lesions, 30 with normal mucosa). Clinical photographs and basic epidemi…

Truvace record1 min read1 sourceEvidence: Moderate · 68/100 — Controlled or validated comparison · n = 350 · 1 independent sourcePeer-reviewedTRV-2026-1166 · cite or check it
Image-Based Diagnosis of Oral Lesions: Performance of a Vision-Language Model versus Human Clinicians
Hospital Universitari Doctor Peset, València 02 by 19Tarrestnom65. CC BY-SA 4.0 · https://creativecommons.org/licenses/by-sa/4.0

In brief

Results AI-Gemini achieved 97.1% accuracy for lesion detection and malignancy classification, with sensitivity 98.5% and specificity 96.2% for malignancy, and 88.0% accuracy for precise histologic diagnosis. The head and neck surgeon achieved the highest accuracy for precise diagnosis (97.7%).

Main points

  1. Objective To evaluate the real-world diagnostic performance of a multimodal large language model (LLM) for image-based assessment of oral mucosal lesions compared with clinicians of varying expertise.
  2. Study design Prospective international multicenter diagnostic accuracy study.
  3. Setting Twenty university and tertiary head and neck centers in Italy, Belgium, France, Spain, and Israel.

The problem

The AI chatbot assigned the correct urgency in only 70% of cases and tended to overestimate risk.

The full story

Objective To evaluate the real-world diagnostic performance of a multimodal large language model (LLM) for image-based assessment of oral mucosal lesions compared with clinicians of varying expertise. Study design Prospective international multicenter diagnostic accuracy study.

Setting Twenty university and tertiary head and neck centers in Italy, Belgium, France, Spain, and Israel. Methods We enrolled 350 consecutive patients (320 with oral lesions, 30 with normal mucosa).

Sources

  1. Peer-reviewedOtolaryngology–Head and Neck Surgery2026-09-21

The debate