Healthcare-Diagnostics-and-Treatment-Recommendation-System

🧠 SmartMed AI

Welcome to SmartMed AI — an AI-powered healthcare assistant that uses multiple intelligent agents to generate preliminary medical diagnoses and personalized treatment plans based on patient-provided information. This interactive web application is built using modern AI frameworks and can be accessed remotely through an automatically generated public URL using Ngrok.

⚠️ Disclaimer: This AI-generated diagnosis and treatment plan is for informational purposes only. Always consult a qualified healthcare professional for actual medical advice, diagnosis, or treatment.

---

📚 Table of Contents

• Overview
• Features
• Technologies Used
• Installation
• Usage
• Public Sharing via Ngrok
• Configuration
• Examples
• Troubleshooting
• What is a Multi-Agent System?
• Contributors

---

🌟 Overview

SmartMed AI allows users to enter basic patient details, such as age, gender, symptoms, and medical history. Once submitted, the system uses a pair of specialized AI agents — one for diagnosis and the other for treatment — to generate intelligent medical insights in real time.

The application can also be shared live over the internet using Ngrok, so that others can interact with the AI from anywhere without any deployment setup.

---

🚀 Features

• 👤 Collects patient demographics and health inputs
• 🧠 Uses multiple AI agents for diagnosis and treatment
• 📄 Provides ranked diagnoses with supporting evidence
• 💊 Suggests treatment plans with medications, lifestyle advice, and follow-ups
• 🌐 Real-time internet access using search and scraping tools
• 🔗 Shareable live app via Ngrok for remote testing or demos

---

🛠 Technologies Used

Technology	Purpose
Streamlit	Builds the user interface
CrewAI	Defines and manages multi-agent workflows
LangChain + OpenAI	Powers the LLM reasoning (GPT-4 / GPT-4O-MINI)
SerperDevTool	Integrates web search capabilities
ScrapeWebsiteTool	Enables real-time web scraping for additional data
python-dotenv	Manages API keys and sensitive config values securely
Ngrok	Publishes the app to the internet for external access
Google Colab	Sample platform used for experimentation

---

🧩 Installation

1. Clone the Repository:

   git clone https://github.com/Naveenpandey27/Healthcare-Diagnostics-and-Treatment-Recommendation-System.git
   cd Healthcare-Diagnostics-and-Treatment-Recommendation-System

2. Create a Virtual Environment (optional but recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Required Dependencies:

   pip install -r requirements.txt

4. Set Up Environment Variables: Create a .env file in your project root:

   OPENAI_API_KEY=your_openai_api_key
   SERPER_API_KEY=your_serper_api_key

5. (Optional) Install Ngrok:

   pip install pyngrok

---

💻 Usage

Running the Application Locally

streamlit run main.py

Running the Application and Exposing it with Ngrok

In ngrok_runner.py or in your notebook, use the following:

import os
from threading import Thread
from pyngrok import ngrok

def run_streamlit():
    os.system('streamlit run /content/main.py --server.port 8503')

thread = Thread(target=run_streamlit)
thread.start()

public_url = ngrok.connect(addr='8503', proto='http', bind_tls=True)
print('Streamlit app is alive at :', public_url)

Result:

After a few seconds, the terminal will print a public URL like:

Streamlit app is alive at : https://abcd1234.ngrok.io

You can send this link to anyone for real-time testing.

---

⚙️ Configuration

You can change the model (gpt-4o-mini), temperature, or token limits directly in the script. The .env file must contain valid API keys for:

• OPENAI_API_KEY
• SERPER_API_KEY

---

🧪 Examples

Sample Input:

Gender: Female
Age: 42
Symptoms: shortness of breath, fatigue, swelling in legs
Medical History: hypertension, type 2 diabetes

Sample Output:

• Diagnosis:

  1. Congestive Heart Failure
  2. Chronic Kidney Disease (possible comorbidity)

• Treatment Plan:

  • Start low-sodium diet
  • Diuretic therapy (Furosemide)
  • Cardiology follow-up
  • Lifestyle changes (weight monitoring, fluid intake tracking)

---

🛠 Troubleshooting

Issue	Solution
App doesn’t launch	Check that Streamlit and dependencies are installed
Blank/empty responses	Verify that API keys are valid and not rate-limited
Ngrok URL not working	Ensure Ngrok is installed, check for port conflicts
Delays in response	Allow processing time due to agent collaboration (optimize model)

---

🧠 What is a Multi-Agent System?

A multi-agent system involves several intelligent agents working collaboratively. In this application:

• The Medical Diagnostician Agent evaluates symptoms and medical history to form a ranked list of possible diagnoses.
• The Treatment Advisor Agent uses the diagnoses to formulate a suitable treatment plan, including medications, lifestyle changes, and follow-up care.

This modular, cooperative setup mirrors real-world medical workflows and improves the decision-making process.

---

👥 Contributors

• Naveen Pandey