MedSynapse: Disease Prediction System 🏥

🏆 Project for "Thinking Machine" Competition - IIIT Pune

"Revolutionizing Early Diagnostics with Artificial Intelligence"

---

🌐 Live Access

The MedSynapse Disease Prediction System is deployed and ready for immediate use.

Note: The live version handles all model loading and processing in the cloud via Streamlit Cloud.

---

📽️ Video Demonstration

Experience a full walkthrough of the MedSynapse Disease Prediction System in action:

The video covers MRI analysis, X-Ray detection, and tabular data predictions, including heart disease prediction.

📺 YouTube Reference

For an alternative viewing experience, watch the demonstration on YouTube: ▶️ YouTube Video Demonstration

---

📑 Table of Contents

• Overview
• Problem Statement
• Solution Architecture
• Key Features
• Scientific Methodology
• Tech Stack
• Installation & Setup
• Project structure
• Future Scope
• Contributors

---

🚀 Overview

MedSynapse is an advanced integrated healthcare platform developed for the "Thinking Machine" competition at the Indian Institute of Information Technology (IIIT), Pune.

The system is designed to assist medical professionals and individuals in the early detection of critical diseases. Leveraging the power of Machine Learning and Deep Learning, our application provides instant, accurate risk assessments across two primary diagnostic domains:

1. Multi-disease Prediction Models (Diabetes & Heart Disease)
2. Medical Image Analysis (Chest X-Ray Pneumonia & MRI Brain Tumor detection)

🎯 Problem Statement & Competition Prompts

Early diagnosis is crucial for effective treatment and management of chronic diseases. This project specifically addresses the Diagnostic Tools track of the Thinking Machine competition, focusing on the following suggested problem statements:

1. Multi-disease Prediction Models: Developing robust models for systemic diseases like Diabetes and Heart Disease using clinical diagnostic data.
2. Medical Image Analysis: Utilizing Deep Learning for anomaly detection in medical imaging, specifically Chest X-Ray (Pneumonia) and MRI (Brain Tumor) analysis.

Our goal is to bridge the diagnostic gap by providing a low-cost, AI-driven initial screening tool that addresses accessibility, cost, and time constraints in modern healthcare.

💡 Solution Architecture

Our solution combines three predictive models into a unified, user-friendly interface:

1. Tabular Data Analysis: Using Random Forest and Logistic Regression for numerical health records (Diabetes & Heart).
2. Computer Vision: Using Convolutional Neural Networks (CNNs) for medical imaging (Chest X-Rays).
3. Interactive UI: A Streamlit-based web app for seamless user interaction.

---

🌟 Key Features & Diagnostic Modules

Module	Purpose	Model / Technique	Key Inputs
🩸 Diabetes	Likelihood Prediction	Random Forest Classifier	Glucose, BMI, Insulin, Age
❤️ Heart Disease	Cardiovascular Risk	Logistic Regression	Chest Pain Type, Max HR, ECG
🩻 Pneumonia	X-Ray Image Detection	CNN (Custom Architecture)	Chest X-Ray (JPEG/PNG)
🧠 Brain Tumor	MRI Scan Classification	Xception (Transfer Learning)	Brain MRI (4 Classes)

---

🔬 Scientific Methodology

MedSynapse follows a rigorous data processing and modeling pipeline:

• Numerical Data: Uses robust scaling and feature engineering to ensure ~98% accuracy in diabetes detection.
• Image Data: Utilizes Transfer Learning (Xception) and Data Augmentation to identify subtle anomalies in MRI and X-Ray scans.
• Validation: All models are cross-validated to ensure generalizability across different patient demographics.

---

🛠️ Tech Stack

Component	Technologies
Frontend
ML Models
Deep Learning	Transfer Learning (Xception)
Environment

---

---

💻 Local Developer Setup

If you wish to contribute or run the system locally:

1️⃣ Clone the Repository

git clone https://github.com/ShivamMaurya14/Disease_Prediction_System.git
cd Disease_Prediction_System

2️⃣ Install Dependencies

pip install -r requirements.txt

3️⃣ Launch the Application

streamlit run app.py

---

📂 Project Structure

├── app.py                # Main Streamlit Application
├── models/               # Trained ML/DL Models (.pkl, .h5)
├── notebooks/            # Jupyter Notebooks for training
├── scripts/              # Helper Scripts (verification, utilities)
├── assets/               # Static assets (images, demo video)
├── requirements.txt      # Python Dependencies
└── README.md             # Project Documentation

---

🧪 Model Verification

To verify that all models are present and loading correctly without starting the UI, run our verification script:

python scripts/verify_models.py

---

� Future Scope

• Mobile App: Develop a Flutter/React Native version for on-the-go access.
• More Diseases: Add modules for Skin Cancer (Dermatology), Kidney Disease, and Liver Disease.
• Doctor Connect: Feature to book appointments with specialists if high risk is detected.
• Report Generation: Download a PDF report of the analysis.

---

👥 Contributors

Team - MEDSYNAPSE

Name	Role	Profile
Shivam Maurya	AI & Robotics Engineer

---

Made with ❤️ for healthcare innovation.