Oropharyngeal Cancer Diagnosis Using Deep Learning Technique with CNN

Abstract

A cancer with a high severity that is widespread and complex is oral cancer. With 130,000 fatalities each year, oral cancer ranks as the eighth most frequent cancer worldwide in India. Salivary glands, tonsils,mouth and  the neck are among areas where the tumour can be found. For example, a biopsy involves taking a small tissue sample from a body component and examining it under a microscope. There are also other screening techniques. However, the drawback is that it is impossible to distinguish between cancer cells and normal cells, and it is also impossible to categorise how many cells are impacted. In this study, cancer cells will be found and categorised in the oral region using digital processing technology. The most effective deep learning method was tested against a histopathological and real-time dataset.  VGG16,ResNet50,MobileNetV2, DenseNet, and VGG19 were the 5 deep learning models utilised in this study. Classification Accuracy, Recall Rate,Error Rate, and Precision Rate are used to compare the effectiveness of the suggested approach. The results of the evaluation showed that ABC, FPSO, and CNN work better together to identify oral cancer.