Deep Learning Based Colorectal Cancer (CRC) Tumors Prediction

Author:- Kamrul Hasan Talukder, Rahul Deb Mohalder
Category:- Conference; Year:- 2021
Discipline:- Computer Science & Engineering Discipline
School:- Science, Engineering & Technology School


Cancer detection and prediction using computer assisted systems has become the most leading research area in recent times. It has a big demand in the medical sector for identifying not only cancer but also any diseases detected and predicted from pathological data or images. Colorectal Cancer or Colon Cancer (CRC) detection is also one of them. Because CRC has become a global health issue day by day. In this paper we used a dataset of 10,000 histopathological images with the same dimension of colonic tissue. We used ensemble methods and classifiers for classifying images. We obtained the best accuracy 99% from XGBoost classifier and from others were 98%, 97%, 96%, 92%, 92% and 89% which exactly classifying 523 colon adenocarcinoma images and 477 benign colonic tissue images from 1,000 histopathological images.

Read More