Neural Networks in Classifying Medical Imaging: Liver Cancer

Abstract

Recent surveys show an increase in the rate of number of people dying due to late diagnosis of liver cancer. The cancerous cells need to be diagnosed in an efficient manner having high accuracy rate. Medical Image Techniques play a vital role in detection and diagnosis of cancer. In this proposed work, we have implemented an Artificial Neural Network (ANN) for detecting liver cancer at an earlier stage which is generally used in various applications for data classification and pattern recognition. Computed tomography (CT) images of the liver are taken as input and then passed through the image pre-processing stage. Feature extraction is done on segmented tumor images and is successfully used to classify the benign and malignant images from our input images. The proposed work attempts to find various classifiers that could classify cancerous images with high accuracy by evaluating its performance on the input images.