Oral Cancer Diagnosis using Deep Learning for Early Detection

Abstract

Oral cancer's overall prognosis is still bad. Over half of those who are afflicted are in advanced stages. Screening testing results and earlier detection strategies for oral malignancies, as previously said, rely in great part on medical professionals' health centre experience and there is no established strategy. The goal of the proposed model is to provide a fast, non-invasive, cost-effective, and simple Deep learning approach to determine oral cavity squamous cell carcinoma (OCSCC) sufferers from photographs. Oral cancers are broad and complex malignancies with a high mortality rate. Oral cancer is the eighth most frequent cancer in India, accounting for 130,000 fatalities per year. The tumour can be found in the glands, tonsils, neck, face, and mouth. Oral cancer can be diagnosed using a variety of ways, including biopsy, which involves taking a small sample of tissue from a section of the body and examining it under a microscope, as well as some screening measures. However, because it is difficult to clearly distinguish cancer cells that have been impacted by cancer [1], cancer cells that have been affected in the oral area are detected and classified using digital processing technology in this study. For early diagnosis and categorization, modern technologies and a deep learning algorithm can be used. The fuzzy particle swarm optimization (FPSO) is used to select the optimal texture characteristic after extracting the texture characteristics. Finally, the Convolution Neural Network classifiers was used to classify the best qualities [3]. Recall Rate, Accuracy Rate, Precision Rate, and Failure Rate are used to compare the performance of the proposed method. The suggested model outperforms existing models in detecting oral cancer, according to the results of the evaluation.