Detection of oral cancer lesions using hybrid classifier

Oral cancer is a main worldwide health problem accounting for 606,520 deaths in 2020, and it is most predominant in the middle-and low-income nations. Permitting computerization in identifying potentially malignant and malignant lesions in the oral cavity would possibly result in low-expense and early detection of the disease. The most important purpose of this research is to find the Oral Cancer Lesions affected region in the tongue images. The current work utilized the GVF algorithm to detect Oral Cancer Lesions using features involved in tongue images. This article offers a novel approach to merging bounding box annotations from different medical practitioners. Additionally, gradient vector flow was used to segment images, where the complicated patterns have been obtained for tackling this difficult task. Using the initial data gathered in this study, a hybrid classifier algorithm was assessed to detect Oral cancer lesions, and features like colour, texture and geometry were extracted. BioMed Chinese Medicine Repository collects the tongue images. Additionally, performances are described categorizing as per the kind of referral decision. Our initial findings establish support vector machine has the probability of challenging this stimulating task.