Oral Health and Maxillofacial Research

ABSTRACT

Oral cancer (OC), the most common malignancy of the head and neck region, remains a global health concern due to its poor survival rates, largely resulting from delayed detection. While histopathology is the diagnostic gold standard, it is prone to inter-observer variability and is time-consuming, contributing to treatment delays. Emerging advancements in artificial intelligence (AI) offer transformative opportunities to address these limitations by enhancing diagnostic accuracy, efficiency, and reliability. This review explores the critical role of AI in oral cancer detection, with a focus on its most promising applications. AI technologies, including machine learning (ML) and deep learning (DL), have demonstrated exceptional capabilities in analysing medical images, such as histopathological slides, radiographs, and intraoral photographs. Deep learning models, particularly convolutional neural networks (CNNs), excel in detecting malignant lesions and precancerous conditions with high sensitivity and specificity. Additionally, AI-powered systems can automate cytological analysis, extract radiomic features for early lesion identification, and support personalized care through risk stratification and prognostic predictions. Telemedicine applications leveraging AI enable real-time diagnostics in underserved regions, improving access to quality care. However, several barriers hinder the widespread adoption of AI in clinical practice. Challenges include the scarcity of high-quality, annotated datasets for model training, the need for seamless integration into clinical workflows, and compliance with ethical and regulatory standards. Furthermore, successful implementation demands clinician training and patient-centred development to ensure usability and trust. This review underscores AI's transformative potential in oral cancer diagnostics while addressing the challenges that need resolution. Future priorities should include fostering data-sharing initiatives, advancing algorithm adaptability for diverse clinical scenarios, and conducting rigorous clinical trials to validate AI tools. These efforts can drive earlier detection and improve outcomes for patients with oral cancer.

KEYWORDS

1. CBCT
2. Oral cancer
3. AI
4. Convolutional Neural Networks
5. Telemedicine