Trends in the Diagnosis and Management of Liver Carcinoma: Advances in Deep Learning Models

Abstract:

Liver cancer is a significant health problem all over the world. Liver lesions are defined as areas of the liver that have been damaged by injury or illness, and their precise location can be pinpointed with the help of diagnostic radiography. Liver lesions refer to abnormal tissues that can be detected in the liver. In this review, the authors summarise the techniques currently used to analyse liver tumours and detect their presence. Relevant articles published over the past 10 years were retrieved from databases such as ScienceDirect, Scopus, PubMed, and Google Scholar. Based on experimental results and comparisons with other methods, the researcher proposes using a deep learning technique, such as a probabilistic neural network, to diagnose and manage liver cancer. This review explores various deep learning-based approaches for liver lesion segmentation. The analysed studies focus on the application of deep learning models for managing liver carcinoma. However, several technical challenges remain to be addressed before computer-aided methods of liver lesion segmentation can be considered fully mature.