Liver Disease Detection Using Machine Learning

Project Overview:

• Feature Selection Using Genetic Algorithms: Identified crucial features that significantly impact liver disease prediction, enhancing the predictive models’ accuracy.
• Machine Learning Integration: ntegrated the selected features into various machine learning models like k-Nearest Neighbors, Random Forest, and Support Vector Machines, further refining the accuracy of liver disease predictions.
• Full Stack Web Application: Developed a user-friendly web application using Flask, which serves as an interface for the predictive models. The application includes multiple pages such as Sign Up, Login, Profile, Prediction Form, and a Landing Page, providing a comprehensive tool for both medical professionals and patients.

Technical Architecture:

• Our methods achieved high accuracy in detecting liver disease, making the tool a valuable adjunct in medical diagnostics.
• Employed a layered architecture approach, facilitating maintenance and scalability.

Results:

• The application architecture supports robust data handling and model deployment, ensuring reliable predictions and a seamless user experience.
• The findings and methodologies were well-received in the medical community, as detailed in our publication in the ‘Annals of the Romanian Society for Cell Biology’.

Download paper here

This project not only pushes forward the boundaries of medical data analysis but also provides a practical tool that could potentially save lives through early disease detection.