Inspiration

Our inspiration came from observing how confusing it can be for everyday users to interpret medical information. From unreadable prescription labels to uncertainty about symptoms — we realized that a simple AI assistant could make a meaningful impact.

We wanted to build something that doesn’t replace doctors, but empowers people with preliminary knowledge and awareness about their health.

What it does

MediScan AI is an all-in-one AI-powered health assistant that helps users analyze medicines, check symptoms, and access first-aid guidance instantly. It leverages Gemini AI’s multimodal capabilities to interpret both text and images, making healthcare more accessible, accurate, and user-friendly.

Users can: • Scan medicines to identify their name, dosage, uses, and side effects. • Check symptoms to understand possible causes and recommended next steps. • Access step-by-step first-aid instructions for common emergencies.

MediScan AI serves as a 24/7 digital health companion, bridging the gap between medical awareness and accessibility.

How we built it

We built MediScan AI using: • Frontend: Vite, React, and Tailwind CSS for a fast, responsive interface. • AI Backend: Gemini API for both text and image analysis. • OCR Module: Extracts and interprets medicine text from uploaded images. • Data Handling: Structured JSON output for consistent and clear responses.

The workflow: 1. The user uploads a medicine image or describes symptoms. 2. Gemini processes the input through text and vision models. 3. The system returns structured insights, including medicine details, symptom analysis, or first-aid steps.

Challenges we ran into

• Ensuring OCR accuracy with low-quality or blurry images. • Managing response latency during image analysis. • Handling limited access to verified medical databases. • Balancing helpful AI responses with ethical boundaries in healthcare.

Accomplishments that we're proud of

• Developed a fully functional multimodal AI healthcare assistant within hackathon constraints. • Integrated Gemini’s text and vision capabilities effectively. • Designed a clean, accessible, and user-friendly interface. • Built a scalable architecture ready for future healthcare integrations.

What we learned

• How to integrate text and image AI models into a single application.
• How to handle OCR and structured data parsing efficiently.
• The importance of designing intuitive and clear health-related user interfaces.
• The ethical considerations and limitations of using AI in healthcare contexts.

What's next for Mediscan-Ai

We plan to expand MediScan AI into a complete AI healthcare ecosystem with: • Voice assistant capabilities for hands-free use. • Health reminders and medication tracking. • User health history for personalized insights. • Integration with wearable devices for real-time data analysis. • Multilingual support for broader accessibility. • A verified “Doctor Connect” feature for professional consultations.

MediScan AI aims to make smart, safe, and accessible healthcare available to everyone.

Built With

Share this project:

Updates