Artificial Intelligence Helps Diagnose Oral Potentially Malignant Disorders: A Systematic Review and Meta-Analysis.

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.