Liver Function Analysis Using Python

Members

• G. Venkata Sai Ram (22501A0557)
• Aakash Kodali (22501A0501)
• G. Rani (22501A0562)
• B.T. Thanusri (22501A0528)
• A. Mrudula (23505A0501)

🥅 Goals of the Project

• Develop a predictive model to identify potential liver dysfunction using machine learning.
• Utilize real-world medical data to analyze liver function parameters effectively.
• Improve diagnostic accuracy and enable early intervention in liver-related disorders.
• Gain experience in using Python libraries for data science and healthcare analytics.

🔄 Process

1. Data Collection: Acquired a liver patient dataset from Kaggle containing 583 records with 10 medical features.

2. Data Preprocessing:

   • Handled missing values (e.g., imputation in the Albumin_and_Globulin_Ratio column).
   • Encoded categorical values such as Gender.
   • Scaled features using StandardScaler for model efficiency.

3. Model Building:

   • Used Logistic Regression for binary classification.
   • Trained the model with stratified train-test split for balanced class representation.

4. Evaluation:

   • Evaluated using Accuracy, Precision, Recall, F1-Score, AUC-ROC, and Confusion Matrix.

5. Result:

   • Achieved reliable prediction accuracy with potential to enhance via advanced algorithms in future iterations.

🧩 Details of Features

• Age: Patient age

• Gender: Male/Female

• Total_Bilirubin: Amount of bilirubin in blood

• Direct_Bilirubin: Conjugated bilirubin level

• Alkaline_Phosphotase: Enzyme level, important liver function indicator

• Alamine_Aminotransferase (ALT): Enzyme related to liver inflammation

• Aspartate_Aminotransferase (AST): Enzyme, often used in liver tests

• Total_Protiens: Protein level in the blood

• Albumin: Main protein in blood plasma

• Albumin_and_Globulin_Ratio: Protein ratio indicating liver health

• Dataset (Target):

  • 1: No liver dysfunction
  • 2: Possible liver dysfunction