Liver Cirrhosis Stage Prediction

Prediction of the stage of liver cirrhosis of a patient.

Problem Definition

Cirrhosis results from prolonged liver damage, leading to extensive scarring, often due to conditions like hepatitis or chronic alcohol consumption.

In this study we'll tyr to predict the stage of the cirrhosis for a person with the given parameters. Since the number of outcomes are limited, we'll approach this problem as a multi-class classification problem. We'll select and train a machine learning algorithm based on the performance metric which will be deciden in the upcoming section.

To install the required libraries, run the following command:

pip install -r requirements.txt

How to run it locally

Terminal

uvicorn main:asgi_app --port 10000

IDE

Create a python run configuration and choose the main.py file

Docker

1. Run docker build -t cirrhosis-prediction .
2. Run docker run -p 10000:10000 cirrhosis-prediction
3. Open your browser and go to http://localhost:10000/

Online

API: https://cirrhosis-prediction-render.onrender.com//predict_api

{
    "N_Days": 1567,
    "Status": "C",
    "Drug": "Placebo",
    "Age": 15373,
    "Ascites": "Y",
    "Hepatomegaly": "N",
    "Spiders": "Y",
    "Bilirubin": 0.3,
    "Cholesterol": 136,
    "Albumin": 3.75,
    "Copper": 15,
    "Alk_Phos": 976,
    "SGOT": 53,
    "Tryglicerides": 78,
    "Platelets": 220,
    "Prothrombin": 10.3
}