MEDI-MINDAI---PROJECT

MediMind AI – Intelligent Prescription Assistant

Understand Your Prescription. Take Control of Your Health.

MediMind AI is a full-stack healthcare web application that helps patients understand prescription text, organize their medicines, and manage daily reminders.

⚠️ Medical safety: MediMind AI is an informational and organization tool. It does not diagnose conditions, prescribe medicines, or recommend dosage changes. It never invents missing prescription details; unclear information is always flagged for manual verification. Always confirm details with your doctor or pharmacist.

ℹ️ Demo note: The OCR (text extraction) and AI explanations in this project are simulated demonstration data. No real OCR or AI service is connected yet.

Features

Landing page with animated prescription mockup, features, how it works, privacy and safety sections
Accounts: create account and sign in (passwords stored as BCrypt hashes), with separate data for each user
Dashboard: statistics, weekly activity chart, today's schedule, recent prescriptions, recent activity
Prescription scanner: drag and drop, browse, mobile camera capture, file type and size validation, image preview, animated four-stage scan
AI analysis (simulated): extracted text, highlighted uncertain handwriting, simplified explanations, medical term definitions
Medicine verification: editable form, unclear fields highlighted, manual confirmation required before saving
My Medicines: add, edit, delete, search, filter, details view
Reminders: daily timeline, mark as taken or skipped, weekly calendar, completion progress
Prescription history: search, filter, view, delete
AI assistant chat: floating chat drawer with scripted answers that only explain general terminology
Settings: profile, notification preferences, privacy, light/dark theme
Responsive design with accessibility support and reduced-motion support

Tech Stack

Layer	Technology
Frontend	HTML5, CSS3, JavaScript (ES6+), single file `index.html`
Backend	Java 17, Spring Boot 3 (Web, Data JPA, Validation)
Database	MySQL
Security	BCrypt password hashing (`spring-security-crypto`), token-based sessions
Build tool	Maven

Project Structure

demo/
├── pom.xml
└── src/main/
    ├── java/com/example/demo/
    │   └── MediMindApplication.java      # entities, repositories, REST controllers
    └── resources/
        ├── application.properties        # database and server configuration
        └── static/
            └── index.html                # complete frontend (served by Spring Boot)

Getting Started

Prerequisites

JDK 17 or higher
Maven (or use the included mvnw wrapper)
MySQL Server running locally

Setup

Clone the repository

git clone https://github.com/KhadijaMulla/MEDI-MINDAI---PROJECT.git
cd MEDI-MINDAI---PROJECT

Configure the database in src/main/resources/application.properties:

spring.datasource.url=jdbc:mysql://localhost:3306/medimind?createDatabaseIfNotExist=true&useSSL=false&allowPublicKeyRetrieval=true&serverTimezone=UTC
spring.datasource.username=root
spring.datasource.password=YOUR_MYSQL_PASSWORD

The medimind database and its tables are created automatically on first run.

Run the application
```
# Windows
.\mvnw.cmd spring-boot:run
```

# macOS / Linux ./mvnw spring-boot:run


4. **Open the app** at `http://localhost:8080` (or the port set by `server.port` in `application.properties`).

### Try it out
1. Click **Get Started** and create an account.
2. Open **Prescription Scanner** and upload a sample prescription image (JPG, PNG or WEBP, max 5 MB).
3. Click **Scan Prescription** and review the simulated results.
4. Verify each medicine. Fill in any highlighted unclear fields and tick the confirmation box.
5. See your medicines in **My Medicines** and your reminders in **Reminders**.

## REST API

| Method | Endpoint | Description |
|---|---|---|
| POST | `/api/auth/register` | Create an account |
| POST | `/api/auth/login` | Sign in |
| GET / PUT | `/api/state` | Load or save the signed-in user's data (requires `Authorization: Bearer <token>`) |
| GET, POST | `/api/medicines` | List or create medicines (`?q=` to search) |
| GET, PUT, DELETE | `/api/medicines/{id}` | Read, update or delete a medicine |
| GET, POST | `/api/reminders` | List or create reminders |
| PUT, DELETE | `/api/reminders/{id}` | Update or delete a reminder |
| PATCH | `/api/reminders/{id}/status?value=taken` | Mark as taken, skipped or pending |
| GET | `/api/history` | Prescription history |
| GET | `/api/stats` | Dashboard statistics |
| POST | `/api/scan` | Upload an image (returns simulated OCR and AI data) |
| POST | `/api/chat` | Scripted terminology assistant |

## Medical Safety Design

- Strength and instructions are shown **only when present** in the extracted text; missing values are never guessed
- Unclear fields are highlighted and must be reviewed by the user
- A medicine cannot be saved without manual confirmation
- The assistant refuses diagnosis and dosage questions
- All AI-style output is clearly labelled as simulated, with a visible warning to verify with a doctor or pharmacist

## Limitations

- OCR and AI responses are simulated, and the scan always returns the same demo result
- Browser-based reminders are a demonstration and are **not** a substitute for a reliable medical alert system
- Authentication is demo-level (no token expiry or HTTPS); not suitable for real patient data

## Future Scope

- Real OCR (for example Tesseract) and a real AI explanation service
- Push, SMS and email reminders
- Doctor and pharmacist sharing, multi-language support
- Production security (Spring Security, JWT, HTTPS) and separate database tables per feature
- Drug interaction information from trusted sources

## Author

**Khadija Mulla**

---

*This project is a hackathon demonstration and does not provide medical advice.*