Convolution Neural Network Model with Feature Linked Vector for Oral Cancer Detection

One of the most common types of cancer, oral cancer is notorious for its high mortality rate due to its tendency to be diagnosed late. Cancers of a lips, tongue, cheeks, the floor of the mouth, tough as well as taste buds, nostrils, and glottis are all considered types of oral cancer. It is impossible to stop the uncontrollable multiplication of cancerous cells in the mouth, and the resulting mutations have the potential to spread to nearby organs. Low-cost and early disease diagnosis may be possible through the detection of vigilant tumors in the oral fissure. We’ve used deep ML algorithms for diagnosis of oral cancer. The information gathered in this study is used to evaluate deep learning-based machine vision strategies for detecting and classifying oral lesions in the interest of early cancer detection. To detect oral cancer in photographs, we trained a deep learning algorithm on a cascade of Convolution neural networks. High detection accuracy for oral cancer compared to earlier models.

I. Introduction

Machine learning is a field of study that enables computers to acquire new skills without being explicitly programmed to do so. One of the most fascinating techniques that has ever built is machine learning [17]. To give the computer a much more humanlike quality, the name suggests that it should be able to learn. Neural nets, which are used in Deep Learning algorithms, are nothing more than a collection of decision-making channels that have been pre-trained to perform a specific task. Every one of these is then processed by successively more simplistic representations. Deep learning methods are mainly used it for categorization and object recognition. When it comes to classification, CNN is a top-tier deep learning algorithm. There have been many iterations of CNN, each with their own set of improvements. After CNN R-CNN algorithm was suggested which was an innovation of CNN algorithm.