Augmentation of Statistical Features in Cytopathology Towards Computer Aided Diagnosis of Oral PrecancerlCancer

Abstract

Oral cancer is a leading malignancy and a rising concern in India. Early detection of the disease is essential at reducing mortality. In this paper, we propose a computer assisted method for diagnosis of oral pre-cancer/cancer using oral exfoliative cytology. A combination of features were extracted from expert delineated cells and nuclei collected from cytology of patients suffering from oral sub-mucous fibrosis, oral leukoplakia or oral squamous cell carcinoma and subject with no lesion. These features were used to train predictive machine learning models like support vector machine, k nearest neighbor, random forest, etc. These models were verified using validation data set. The verification experiments showed promising results with the random forest classifier having a test accuracy of 90%.