Intelligent Hepatocellular Carcinoma Detection using Hybrid Radial basis Function Networks

Patients with hepatocellular carcinoma may have a higher chance of survival with earlier detection and treatment. It takes a lot of time and effort to manually identify malignant cells. In order to make decisions that result in an efficient treatment, diagnosis methods based on computational techniques is essential. In this work, deep learning approaches are employed to automatically determine feature representation from sample images using radial basis function networks, as opposed to traditional image-dependent “contextual” feature evaluation by professionals. The primary goal of the proposed model is to detect hepatocellular carcinoma in diagnostic imaging. The suggested model accomplishes this by using pre-processing in two steps, namely image denoising using a Wiener filter and segmentation using region-based segmentation technique. The Radial Basis Function Networks model’s hyper parameters are tweaked using the Salp Swarm optimization Algorithm technique, which is then used to perform the classification of hepatocellular carcinoma. The proposed model showed superior performance compared to the existing methods.