Artificial Intelligence in Early Detection and Diagnosis of Oral Cancer: A Machine Learning-Based Clinical Evaluation

Abstract

Oral cancer remains one of the most challenging malignancies worldwide due to its high mortality rate, delayed diagnosis, and significant impact on patients' quality of life. Early detection is the most effective strategy for improving survival outcomes, yet conventional diagnostic approaches rely heavily on clinical expertise, biopsy procedures, and histopathological examination, which may lead to delayed intervention, especially in resource-limited settings. Recent developments in artificial intelligence (AI), particularly machine learning (ML) and deep learning (DL), have introduced innovative opportunities for automated detection and diagnosis of oral cancer using clinical images, histopathological slides, radiographic imaging, and patient health records. This study presents a machine learning-based clinical evaluation framework designed to assess the effectiveness of AI in the early detection and diagnosis of oral cancer. A retrospective dataset consisting of 5,240 clinical oral images and histopathological records collected from three tertiary healthcare institutions was utilized for model development and evaluation. Multiple supervised learning algorithms, including Support Vector Machine (SVM), Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Convolutional Neural Networks (CNN), were comparatively analyzed using standardized preprocessing techniques and five-fold cross-validation. Performance evaluation was conducted using accuracy, sensitivity, specificity, precision, F1-score, and Area Under the Receiver Operating Characteristic Curve (AUC).