Bridging Healthcare Gaps: Sustainable Liver Disease Diagnosis Using Machine Learning in Resource-Limited Regions

Liver disease is a major global health issue, and proper diagnosis at the right time is crucial for treatment. This paper discusses machine learning methodologies in the detection and diagnosis of liver diseases such as hepatitis, cirrhosis, and hepatocellular carcinoma. It addresses the diagnostic complexities such as nuances in liver function tests and variability in symptoms and explores how ML can leverage diverse data sources, including medical imaging, genetic markers, and clinical data, to improve precision and efficiency. The study applied advanced Python tools and algorithms with 10-fold cross-validation for splitting the dataset. Random Forest had the highest accuracy at 99%, followed by SVM and Gaussian Naïve Bayes. The research will provide an affordable diagnostic tool, especially for developing countries with a high burden of liver disease, which will contribute to SDGs 3, 1, 10, and 9 through improved access to healthcare, reduction of economic burdens, and innovation. The future directions are multimodal data integration and personalized diagnostic systems for further improving the outcomes in patients.