Design of Novel Hybrid Model for Detection of Liver Cancer

Abstract

Liver cancer being one of the most severe categories of cancerous diseases has lead to the death of numerous people. Images of Liver tumor by Computed Tomography (CT scan) have much noise in it, because of which diagnoses of tumor level becomes difficult. Changes in anatomy differ for different patients. This makes automatic recognition of tumor using CT scan images very challenging. To overcome this problem, in this paper, detection of liver cancer is done using fused images of CT scan and Magnetic Resonance Imaging (MRI) that plays a crucial role in the choice of different strategies for liver disease and also for treatment monitoring. This research article proposes a design for early detection and classification of liver cancer which consists of discrete wavelet transform image fusion technique, extraction of Speeding up robust features, feature selection using Cuckoo meta-heuristic approach and machine learning algorithms from contrast enhanced CT and MRI images. The proposed framework is fully automated which requires no user interaction.