Medical Test Result Prediction

This project is an end-to-end medical AI application that predicts patient test results based on structured clinical data.

Features

• Machine Learning model (ensemble) trained on real medical dataset
• Predict test result: Abnormal, Normal, or Inconclusive
• Streamlit-based Web UI for interactive usage
• FastAPI backend for programmatic access
• Visual confidence (probability) bar chart
• Encoders & scaler included for inference

How to Run the App

Step into the app folder

cd app

Run FastAPI backend

fastapi run main.py

Visit: http://127.0.0.1:8000/docs to test the API.

---

Run Streamlit Web App

streamlit run app.py

Visit: http://localhost:8501 to use the web UI.

Web application

---

Next step

• Deploy on Docker
• Deploy on AWS
• Model monitoring and MLOps building

---

Data source: Kaggle Healthcare Dataset