Predictive AI Healthcare Insighter

Demo Video: Demo
Privacy Policy: View Here
Terms of Service: View Here

⚠ Note: The hosted URLs may become unavailable after the hackathon results are announced.

---

📌 Overview

What is this app?
Predictive AI Healthcare Insighter is an AI-powered decision-support tool integrated with Electronic Health Records (EHR) via MeldRx.

What does it do?

• Provides real-time patient risk predictions.
• Suggests AI-powered recommended actions.
• Offers preventive care recommendations.
• Displays interactive visual insights about patient history and trends.

---

🔍 Features

✔️ Secure Authentication – Uses OAuth 2.0 authentication via MeldRx.
✔️ AI-Driven Insights – Uses Gemini AI for predictive analytics.
✔️ Risk Score Calculation – Determines the likelihood of disease progression.
✔️ Treatment Recommendations – Provides AI-suggested treatments or preventive care actions.
✔️ Data Visualization – Graphs, charts, and trends displayed via Recharts.
✔️ CDS Hook Integration – Embeds within EHR, showing AI insights when patient data is accessed.
✔️ PDF Reports – Generates downloadable AI analysis reports for reference.

---

🏗️ Technologies Used

This project is built with:

Technology	Purpose
oidc-client-ts	Authentication & Authorization
Redux Toolkit	Global State Management
React Redux	State Management
Axios	API Requests
Recharts	Data Visualization
TailwindCSS & DaisyUI	UI Styling
Google Generative AI	AI-driven Predictions
React PDF Renderer	PDF Generation

---

🚀 Installation & Setup

Follow these steps to set up the project locally.

1️⃣ Clone the Repository

git clone https://github.com/JonathanSolvesProblems/MeldRX-PredictiveAI-Application.git
cd predictive-ai-service

2️⃣ Create Environment Variables

Create a .env.local file in the root directory and add the following:

NEXT_PUBLIC_AUTHORITY=https://app.meldrx.com
NEXT_PUBLIC_CLIENT_ID=<meldrx client id>
NEXT_PUBLIC_CALLBACK_URL=http://localhost:3000/callback
GEMINI_API_KEY=<Google Gemini API Key>
NEXT_PUBLIC_APP_ID=<meldrx bundle id>

3️⃣ MeldRx App Setup

To integrate with MeldRx, follow these steps:

1. Create an app on MeldRx.
2. Set the required OAuth Scopes for authentication:
   • profile
   • openid
   • launch
   • launch/patient
   • patient/*.*
   • patient/*.read
3. Configure the Redirect URL:
   • For the deployed app:

     https://meld-rx-predictive-ai-application.vercel.app/callback

   • For local development:

     http://localhost:3000/callback

4️⃣ CDS Hook Setup

To enable Clinical Decision Support (CDS) Hooks, follow these steps:

1. Navigate to your MeldRx App Dashboard.

2. Locate the CDS Hooks Configuration section.

3. Set the CDS Hook Service URL:

   • For the deployed app:

     https://meld-rx-predictive-ai-application.vercel.app/api/cds-services/0001

   • For local development:

     http://localhost:3000/api/cds-services/0001

     3.1. Ensure to host it with ngrok generated URL and use that as the URL to get it to work without hosting it: ngrok

4. Save the configuration and ensure your CDS Hook is active.

5. The CDS Hook will trigger when a patient record is opened, providing a link to the application for further analysis.

5️⃣ Import Sample Patients

To test the application, you can import sample patient data into MeldRx:

1. Go to the MeldRx Dashboard and navigate to the Patients section.
2. Look for the option to Import Sample Patients and follow the on-screen instructions.
3. Once imported, open a sample patient record within the EHR system.
4. The CDS Hook should trigger automatically, displaying a link to the Predictive AI Healthcare Insighter app.
5. Click the link to access AI-generated insights, including risk scores, recommended actions, and data visualizations.

✅ At this point, the app should be fully functional within the EHR system, displaying AI-driven insights on patient data! 🚀

🌍 Online Deployment

The application is hosted on Vercel at:

👉 MeldRx Predictive AI App

⚠ Note: The hosted URL must be navigated through the MeldRx platform in order to initiate the authentication step to access the application.