Empowering Healthcare with Deep Learning for Oral Cancer Prevention

Oral cancer early detection and prediction are of paramount importance due to the profound impact they have on patient outcomes, healthcare systems, and public health. First and foremost, early detection significantly improves the chances of survival. When oral cancer is diagnosed in its initial stages, treatment options are more effective, less aggressive, and less debilitating. This translates into higher survival rates and a better quality of life for patients. Moreover, early detection reduces healthcare costs. Treating advanced-stage oral cancer requires more resources, including extensive surgeries, radiation therapy, and chemotherapy. Identifying the disease early minimizes the need for such intensive interventions, thereby decreasing the financial burden on healthcare systems and patients. Therefore, in this paper, we have made an attempt to detect and predict oral cancer at an early stage using Deep Learning techniques such as Convolutional Neural Networks with different activation functions and transfer learning methods such as ResNet50, VGG16, VGG19, and AlexNet. It can be seen that the proposed Customized CNN model outperforms by giving an accuracy of 87.6%.