AI Model Detected Oral Lesions With High Accuracy
A new multimodal AI tool showed success in distinguishing malignant from benign oral cavity lesions in a clinical study.
Updated on Sept. 25, 2026 in Cancer
Live Poll
Would you trust an artificial intelligence tool to assist your doctor with a medical diagnosis?
Researchers have developed a multimodal artificial intelligence model that accurately identifies and classifies oral cavity lesions as malignant or benign. This diagnostic tool, tested across 20 centers, aims to assist healthcare providers in triage and early detection of oral cancer.
Why it matters
Early detection of oral cancer is critical for improving patient outcomes, yet distinguishing suspicious lesions from benign ones can be challenging in routine practice. This model may eventually streamline specialist referrals and ensure that high-risk patients receive care more quickly.
In a study involving 350 patients across 20 centers, an AI system based on Gemini 2.5 achieved 97.1% accuracy in detecting malignant lesions compared to histological gold standards. While the tool demonstrated 98.5% sensitivity, independent verification is still required.
The players
University of Sassari
An Italian academic institution that coordinated the multisite research into AI diagnostic models.
The details
The model uses a standardized protocol to analyze images of the oral mucosa to classify tissue health. By comparing these AI-generated findings against histological examinations, researchers have begun to refine the system's ability to distinguish malignant conditions from benign ones. The model also aims to provide an urgency assessment for specialist evaluation, an area where it correctly classified cases 70% of the time.
Timeline
September 21, 2026: The research was published in a scientific magazine.
Health Landscape
This research contributes to the growing field of multimodal AI diagnostic tools designed to assist in territorial triage and telemedicine. It represents a significant effort to bridge the performance gap between general practitioners and sub-specialists in identifying potential malignancies.
While this tool shows promise for clinical use, it is not yet available for personal diagnosis or home screening. If you notice persistent sores, lumps, or unusual patches in your mouth, continue to raise these concerns with your primary care doctor or dentist for a formal evaluation.
The takeaway
Artificial intelligence tools are being developed to improve the accuracy of oral cancer screening and patient triage. Readers should maintain regular dental checkups, which remain the standard for identifying suspicious oral lesions early.
Further reading
For more information on diagnostic advancements and prevention, visit our Cancer hub.
Source note: This article includes information reported by Unica Radio.
Live Poll
Would you trust an artificial intelligence tool to assist your doctor with a medical diagnosis?
