Automated Detection of Oral Cancer and Dental Caries Using Convolutional Neural Network

Dental Health is an important part of our overall health but we often miss to take care of it. Advancement in Artificial Intelligence and Machine Learning has given us immense contributions and one of them is Image Classification. Using Image Classification algorithms now we can easily classify images of things or objects into groups based on their differences like in colour, structure or texture. Same can also be used to identify oral diseases in humans based on the images of their mouth or teeth. Features of the images are used to classify them by comparing them to the dataset of images. A Convolutional Neural Network (CNN) framework was adopted. Out of 598 intraoral images collected, 402 images were used for training and 196 images were used for testing. The model used is a sequential model and it will categorize the images into three classes, “Normal teeth”, “Caries”, “Oral cancer”. The detected disease accuracy on the tested tooth photos was 83-94%. This system helps in early diagnosis of certain oral issues that are typically identified by dentists just by looking at them. It will help dentists to easily diagnose dental-health problems with efficiency and accuracy.