Detection and accurate classification of oral cancer are vital for improving patient outcomes and guiding effective treatment. This study presents a deep learning based framework that integrates image processing for oral cancer classification and localization. The Classification of oral cancer images is done into five classes: No lesion, no referral needed, refer for other reasons, refer–low-risk OPMD, and refer–cancer/high-risk OPMD. Simultaneously, YOLOv2 detects and localizes cancerous regions for precise visualization. Model performance is evaluated using accuracy, precision, recall, and F1-score. Results show that combining ResNet-101 for classification with YOLOv2 for localization enhances diagnostic accuracy and reduces misclassification but the proposed system supports clinicians in early detection of potentially malignant oral disorders and presents an automated system for the detection and classification of oral cancer using a Light Convolutional Neural Network (Light CNN). Oral images are first preprocessed using techniques such as resizing, noise removal, and normalization to enhance image quality. The Light CNN model is then used to extract important features and classify the images into normal and cancerous categories. Compared to deeper networks, Light CNN reduces computational complexity, training time, and memory usage while maintaining high accuracy.
Babu et al. (Thu,) studied this question.