ABSTRACT
The Oral Squamous Cell Carcinoma (OSCC) is a common oral cancer which damages the oral epithelium. The sixth most common cancer in the world is oral cancer. Approximately 4% of all cancers are oral cancers which result in 2% of cancer deaths in the world. The golden confirmation is to analyze biopsies (histopathological), backed by the cancer expert, where the expert takes considerable period of time to examine the oral cancer. The experts usually segment the biopsies manually (theoretically in mind) with the help of microscope. Such processes are prone to error. Hence, there is a need for effective Computer Aided Diagnosis (CAD) system which can help to classify the normal epithelium layer and OSCC effected epithelium layer. Although there are several approaches used to assist the expert for diagnosis of oral cancer yet there are many issues like inaccuracy in the identification and incorrectness in classification of various samples of epithelium layers of oral mucosa, unsuitable preprocessing technique, etc. Thus, an effective approach to segment and classify the histopathological image dataset for identifying cancer in the epithelium layer of oral mucosa with more accuracy has been proposed. This approach has been evaluated and found to provide 74.16% accuracy. It is expected that usage of this approach in the Computer Aided Diagnosis system can help the experts/ doctors to make more accurate decisions in the diagnosis of oral cancer.
