Oral Cancer Detection Using Deep Learning Approach

Oral cancer is a prevalent malignancy that is challenging and has a high severity level. In India, oral cancer is the eighth most prevalent cancer in the world, claiming 130,000 lives annually. The tonsils, salivary glands, cheek, mouth, neck, and tonsils are all affected by the tumor. There are several ways to diagnose oral cancer, including a biopsy, which involves taking a small sample of tissue from a body region. Some screening techniques are also examined under a microscope. The drawback of this method is that it cannot distinguish between cancer cells and classify the amount of cells impacted by cancer, so in this study, digital processing technology will be used to locate and categorize the cancer cells in the mouth area. The utilization of modern technologies and a deep learning algorithm is possible for early detection and categorization. Three feature extraction methods, including wavelet features, Zernike moment, and bagged histogram of oriented gradients, are used in this study. The optimum texture characteristic is chosen using the fuzzy particle swarm optimization (FPSO) technique once the texture characteristics have been loaded. Finally, a convolutional neural network (CNN) classifier was used to categories these traits. Compare the classification accuracy, precision rate, the recall rate, and mistake rate of the proposed method's performance. The evaluation's findings demonstrated that ABC (Artificial Bee Colony) FPSO, and CNN work best together to identify oral cancer.