Liver Cancer Diagnosis with Lightweight Federated Learning Using Identically Distributed Images

A major obstacle for cancer research is the prediction of liver cancer progression. This research looked at models that can predict how Hepatocellular carcinoma (HCC) would develop. As medical science has advanced, records of people with liver cancer at different stages have been collected, and the illness remains the top cause of cancer-related mortality. Keeping private information safe, this study presents a streamlined architecture for federated deep learning to categories liver cancer illnesses. This system uses a distributed client-server architecture to safeguard client data and IID for validating federated deep learning models. The framework was tested with a single client and numerous clients in both conventional and federated learning contexts. We compared several pre-trained models and ultimately settled on AlexNet as our standard because of its exceptionally high accuracy of 98.53 percent. Then, federal learning (FL) was used to assess IID datasets. Due to its enhanced capabilities, the framework is the best choice for diagnosing and classifying liver cancer at an early stage.