Artificial Intelligence Performance in Diagnosing Oral Potentially Malignant Disorders via Intraoral PhotographyArtificial intelligence shows high accuracy detecting oral precancerous lesions

JDR clinical and translational research Published October 8, 2026 Study authors: Horvath D, Fekete A, Martinekova P, Kiss-Dala S, Abram E, Marton K, Zsembery A, Hegyi P, Brody A PubMed ↗ DOI ↗ Editorial oversight: Dr. Lars van Dijk, PhD · Surgical, Procedural & Diagnostic

AI-generated summary of the cited source, checked by automated accuracy review. How we work

Key Takeaway

AI models show high sensitivity and specificity for identifying oral potentially malignant disorders from intraoral photos.

A systematic review and meta-analysis evaluated the diagnostic accuracy of artificial intelligence architectures in identifying oral potentially malignant disorders (OPMDs). The analysis utilized 898 intraoral photographs compared against the gold standard of histological examination to determine the efficacy of automated screening.

The results indicated high diagnostic performance, with a sensitivity of 0.94 and a specificity of 0.95. The diagnostic odds ratio was notably high, and the positive likelihood ratio reached 16.89, suggesting that AI models are highly effective at identifying suspicious lesions from photographic data.

Clinicians can utilize these AI tools to improve screening workflows. High negative predictive values may help identify low-risk cases, potentially reducing the number of unnecessary specialist referrals while ensuring high-risk patients are flagged for further investigation.

While the results are promising, histological examination remains the definitive gold standard for diagnosis. Clinical judgment must remain the primary factor in patient management, and further prospective studies are required to determine the best methods for integrating these tools into routine clinical practice.

How this fits prior evidence

This meta-analysis extends prior coverage of deep learning for OPMD diagnosis by pooling diagnostic accuracy estimates across AI architectures, reporting a sensitivity of 0.94 and specificity of 0.95. It contrasts with the SMMI framework, which is a conceptual risk stratification model rather than a diagnostic tool. The findings also complement evidence on molecular markers and PD-L1, which address risk prediction rather than image-based detection. However, the authors emphasize that histology remains the gold standard and prospective studies are needed.

Detecting early signs of oral cancer is vital for patient safety. A new review looked at how artificial intelligence (AI) performs when analyzing photos of the inside of the mouth to find potentially precancerous conditions. The study looked at 898 images to see if AI could match the accuracy of traditional tissue exams.

The results showed that AI had high sensitivity and specificity, meaning it was very good at correctly identifying both healthy tissue and concerning lesions. Specifically, the AI showed a high negative predictive value. This means that if the AI says a spot is not a concern, it is very likely to be correct. This could help doctors quickly rule out issues and prevent patients from needing unnecessary specialist appointments.

While the results are promising, it is important to remember that a physical tissue exam is still the gold standard for diagnosis. AI is a helpful tool for screening, but it does not replace the judgment of a doctor. More studies are needed to see how these tools can best fit into everyday dental and medical care.

What this means for you:

AI shows high accuracy in identifying oral lesions from photos, helping to rule out issues and reduce unnecessary visits.

Common questions

How accurate is artificial intelligence at identifying oral lesions?

The study found that artificial intelligence had a sensitivity of 0.94 and a specificity of 0.95 when analyzing intraoral photographs. These numbers indicate that the AI was very effective at correctly identifying both healthy tissue and potential issues compared to standard examinations.

Can AI replace a doctor's examination of the mouth?

No, a physical tissue examination is still considered the gold standard for diagnosis. While AI tools have high accuracy and can help identify lesions from photos to reduce unnecessary specialist visits, clinical judgment from a healthcare professional remains essential for patient care.

How does AI help patients with oral health concerns?

Because the AI showed a high negative predictive value, it can help identify which lesions are likely not a concern. This can help patients avoid unnecessary specialist consultations by quickly filtering out cases that do not require advanced intervention.

Study Details

Study typeMeta analysis

EvidenceLevel 1

PublishedOct 2026

PMID41562124

View Original Abstract ↓

INTRODUCTION: Oral potentially malignant disorders (OPMDs) can lead to oral cancer, which is one of the most common cancers worldwide. Prevention is crucial in the avoidance of malignant transformations of OPMDs. Artificial intelligence (AI) provides a new and noninvasive tool for analyzing medical data, such as patient data, radiologic images, and clinical photographs. These AI-based tools can help in the decision-making process. However, histological examination is still the gold standard for diagnosing OPMDs. OBJECTIVES: This systematic review and meta-analysis aimed to investigate the diagnostic accuracy of artificial intelligence on intraoral photographs of patients with OPMDs. METHODS: A systematic search was conducted on 5 major databases (MEDLINE, Embase, Cochrane Library, Scopus, and Web of Science) on November 10, 2023. Included studies compared AI methods to histology examination as the reference. A quantitative analysis was carried out to assess sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), diagnostic odds ratio (DOR), positive likelihood ratio (LR), and negative likelihood ratio (LR) calculated with 95% confidence intervals (CIs). RESULTS: Six eligible articles were included, with 898 images out of 4,046 tested using AI-based architectures. Five studies investigated at least 2 AI models. The overall sensitivity, specificity, DOR, LR, and LR were 0.94 (95% CI, 0.88 to 0.95), 0.95 (95% CI, 0.85 to 0.98), 212.39 (95% CI, 56.39 to 800.00), 16.89 (95% CI, 5.72 to 48.68), and 0.08 (95% CI, 0.05 to 0.13) for the best-performing AI-based architectures in terms of sensitivity, respectively. CONCLUSION: AI-based diagnostic tools have high negative predictive value that could help identify OPMD lesions using intraoral photographs.Knowledge Transfer Statement:This systematic review on AI-based methods to diagnose oral potentially malignant disorders showed that although their high negative predictive value could reduce unnecessary specialist consultations, clinical judgment remains paramount. Further prospective studies are needed to evaluate the integration of AI diagnostics into routine care and screening and policies to enhance efficiency and support early detection and prevention of oral cancer.

Artificial Intelligence Performance in Diagnosing Oral Potentially Malignant Disorders via Intraoral PhotographyArtificial intelligence shows high accuracy detecting oral precancerous lesions

How this fits prior evidence

Common questions

Study Details

Global GJB2 mutation prevalence in nonsyndromic hearing impairment

One in four people with hearing loss carries a specific gene mutation

Clinical research that matters. Delivered to your inbox.

Artificial Intelligence Performance in Diagnosing Oral Potentially Malignant Disorders via Intraoral PhotographyArtificial intelligence shows high accuracy detecting oral precancerous lesions

How this fits prior evidence

Common questions

More on oral potentially malignant disorders

Study Details

Global GJB2 mutation prevalence in nonsyndromic hearing impairment

One in four people with hearing loss carries a specific gene mutation

Clinical research that matters. Delivered to your inbox.

Related in ENT (Otolaryngology)

From Other Specialties