A Framework of Medical Image Segmentation using Edge Stop Function based on Contour Model
W Nancy, Kumar.P Sathish, S. Kirubakaran, Thomas D Ruban, Karthikeyan Mohanraj, R. Azhagumurugan · 2022
Liver cancer is a major taker of lives in recent years. Generally, a liver cancer death rate is very high because the disease shows no symptoms and it's often not caught until the final stages. Traditional Segmentation methods include Region-Based which is based on Intensity and Gradient information which can be applied only when the images has constant gradient information. The Edge-Based Segmentation technique, though can be applied to images with diverse homogeneities, can't be applied to images with non-definitive boundaries. We propose a novel method for liver cancer detection using robust edge stop functions based on Contour detection and Artificial Neural Networks with the Wavelet Features separated from the source input image. By incorporating the gradient information along with the probability scores from a typical classifier, an algorithm is constructed which is applied for the ESF Model. Also the wavelet features classify a dataset and then get the result in normal and abnormal stages. The K-nearest neighbors and SVM approve the effectiveness of the proposed approach.