Development of Hybrid Bi Neural Network for Enhanced Pattern Recognition of Liver Cancer Images

Abstract

newline Segmenting the liver and tumor from abdominal Computed Tomography (CT) images is a challenging task due to the overlap in intensity and variability in the position and shape of soft tissues. Conventional methods relying solely on gray levels or shapes struggle to achieve accurate results. However, the recent advent of deep learning (DL) has shown promising potential in various fields, particularly in medical imaging. This research focuses on the application of DL in liver disease diagnosis, specifically in the areas of liver segmentation, lesion delineation, disease classification, and process optimization. DL offers advantages such as improved accuracy, convenience, safety, and cost-effectiveness compared to traditional pathological biopsies. A unique DL architecture, called the Bi-neural network algorithm (BNN), is presented in this work to enhance the accuracy of liver and tumor classification. The BNN achieves an impressive accuracy of 98.94% in liver and tumor segmentation, outperforming existing algorithms. The outcomes of this study are compared with the results obtained from other current algorithms, validating the superiority of the proposed BNN. The implementation of DL-optimized imaging diagnosis demonstrates its potential to revolutionize liver disease diagnosis and treatment planning. This research contributes to the advancement of medical imaging and liver disease diagnosis through the innovative BNN. The potential for accurate and automated liver and tumor segmentation has significant implications for healthcare, offering a non-invasive and cost-effective alternative to traditional biopsy procedures. newline.