CLASSIFICATION OF COLON CANCER BY USING CNN AND CAPSULE NEURAL NETWORK

Colorectal cancer typically originates as a button-like growth termed a polyp on the surface of the intestinal lining or rectum. The intestine or rectumdivision may invade nearby or adjacent lymph nodes. Due to the fact that blood flows from the intestine’s wall and asubstantial portion of the rectum to the liver, colorectal cancer can metastasize to the liver after spreading to adjacent lymph nodes. Machine Learning obtained a good performance for colon cancer detection. However, the cancer detection systems based on ML need manual detection of the features and separate classifiers for the detection, making the system more complex and time-consuming when using big data. There are several traditional techniques which are not flexible, robust and time consuming as they are devised for manual assessment of colon cancer. Hence, in this research several deep learning techniques namely convolutional neural network (CNN) and Capsule Neural Network are compared. The comparative assessment showed Capsule Neural Performs Better than CNN.