Finding Patterns in Liver Function Test results to interpret Well-defined Liver Diseases
Harshil Kananthoor1, Rajat Kumawat2, Rahul Mohata3, Vaishnavi Bisen4
1Harshil Kananthoor, School Of Engineering And Technology, Dy Patil University Ambi Pune
2Rajat Kumawat, School Of Engineering And Technology, Dy Patil University Ambi Pune
3Rahul Mohata, School Of Engineering And Technology, Dy Patil University Ambi Pune
4Vaishnavi Bisen, School Of Engineering And Technology, Dy Patil University Ambi Pune
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Liver function tests (LFTs) are crucial for diagnosing and monitoring liver diseases, but interpreting their results is often challenging due to the complex patterns associated with various conditions. This study aims to identify distinct LFT result patterns that differentiate specific liver diseases, enhancing diagnostic accuracy. We will analyze a comprehensive dataset of LFT results from patients with hepatitis, cirrhosis, fatty liver disease, and liver cancer. Key parameters include ALT, AST, ALP, GGT, total and direct bilirubin, albumin, and prothrombin time. Advanced statistical and machine learning techniques, such as clustering algorithms, principal component analysis (PCA), and decision trees, will be employed to identify correlations and patterns among these parameters. The findings will be validated using a separate dataset to ensure reliability. We anticipate discovering specific combinations and ranges of LFT parameters that correspond to different liver diseases. These patterns will be presented in a simplified format to aid clinicians in making accurate and timely diagnoses. This research aims to enhance the interpretive value of LFTs, providing a robust framework for diagnosing liver diseases and potentially improving patient outcomes through earlier and more precise treatments.
Key Words: Liver Function Tests, Pattern Recognition, Liver Diseases, Diagnostic Accuracy, Machine Learning, Biomarkers.
