Detection Of Oral Cancer In Hyperspectral Images Using Restricted Boltzmann Machines

Document Type : Primary Research paper


Bannari Amman Institute of Technology, India


Oral cancer is one of the dreadful diseases that affect the people of ages above 40
in most cases. It affects the regions around the mouth especially the back part of mouth
which can lead to death often. There are various compu- tational techniques available to
detect this widespread disease at the later stage only. If the disease is detected at an earlier
stage, then the survival rate of the victims can be increased to 5 years. This paper focuses
on detecting oral can- cerous cells at an earlier stage using deep learning techniques as
they work ex- tremely well for image recognition and image classification. An intelligent
technique comprising of Restricted Boltzmann Machine (RBM) is applied for
differentiating the benign and malignant tissue in hyperspectral images (HSI). After
experimental results, accuracy obtained was 95.75% using the proposed enhanced RBM


Volume 12, Issue 3 - Serial Number 3
ICMMNT-2021 International Virtual Conference on Materials, Manufacturing and Nanotechnology, 30th June, 2021.
June 2021
Pages 1760-1766