An Automated Liver Disease Detection System using Machine Learning and Smart Contract

Abstract

Liver disease is one of the leading reasons of death in both men and women worldwide. Blood pressure, Bilirubin level, Energy level, SGOT & SGPT level, blood sugar level, and Total Protein are all factors that can be observed to predict liver disease in its early stages. The existing healthcare infrastructure is being transformed by technology. With the development of the Internet of Things (IoT), doctors can now remotely monitor and analyze patients, keep their data, and process it for further examination. However, there is a requirement to introduce advanced and improved secured algorithms for quick event processing and detection. In this work, an integrated Smart Contract (SC) and Machine Learning (ML) based system is presented for detecting liver disease that, in addition to storing data in the blockchain, notifies the doctor about the patient's health status. blockchain technology provides a secure environment for storing confidential patient data. Since blockchain data is tamper-resistant, it effectively prevents unauthorized access and data falsification. Using the concept of ML, decision tree classifier variants, such as J48, Hoeffding Tree, Decision Stump, and Random Forest, are applied to the liver patient dataset for efficient classification in order to design the proposed system. The best classifier is chosen based on performance metrics like accuracy, precision, recall, F- score, etc. Finally, the selected classifier's rules are encoded in the smart contract to allow automatic and timely detection of disease. This work also discusses the proposed work's functioning and the evaluation results.