-
-
“Med-Decode: Turning complex medical report into simple English/Tamil/Hindi – Upload Your Report, Get Instant Clarity!
-
MedDecode homepage allowing users to upload medical reports in image or PDF format.
-
Uploading a medical report for AI-based explanation.
-
AI-generated explanation of a medical report in simple, easy-to-understand language without diagnosis.
-
MedDecode also flexible for other language by choosing it and also read the explanation accordingly.
-
Uploading same details in Pdf format to check for conversion
-
MedDecode successfully processing and explaining a PDF medical report using OCR and AI.
Inspiration
Medical reports are often filled with complex terminology that is difficult for non-medical people to understand. This lack of clarity can create confusion and anxiety, which I experienced in my personal life. MedDecode was inspired by the idea of making medical information more accessible and understandable for everyone, without replacing doctors or giving medical advice.
What it does
MedDecode allows users to upload medical reports in the form of images or PDFs. The system extracts the text using OCR and then uses AI to explain the medical terms in simple, understandable language. The platform strictly avoids diagnosis or treatment suggestions and focuses only on explanation.
How we built it
Flask was used to build the web application Tesseract OCR was used to extract text from images Poppler was used to process PDF files ** Google Gemini AI** was used to generate simplified explanations Python handled backend logic and data flow
Challenges we ran into
Handling PDF and image formats reliably Fixing OCR inaccuracies Correctly configuring external tools like Tesseract and Poppler on Windows Ensuring that AI did not provide diagnosis or medical advice Adding of Language other than English.
Accomplishments that we're proud of
Successfully built a fully working AI-powered web application within the hackathon timeline. Implemented OCR to accurately extract text from both medical report images and PDFs. Integrated Google Gemini AI to translate complex medical terminology into simple language. Designed the system to strictly avoid medical diagnosis or treatment advice, ensuring ethical and responsible AI usage. Overcame multiple technical challenges including OCR errors, PDF processing, API configuration, and deployment issues. Created an end-to-end solution that addresses a real-world problem and improves accessibility to medical information.
What we learned
Integrating OCR with AI models in real-world applications Handling file uploads and processing securely Prompt engineering to control AI behavior Building an end-to-end AI-powered web application under time constraints.
What's next for MedDecode -AI powered medical report translator
Multi language support other than tamil,english and Hindi Highlighting abnormal values visually Improved UI/UX User accounts and history
Note: Although a partner challenge selection was required for submission, this project does not directly use ElevenLabs APIs. The project primarily focuses on medical report simplification using OCR and Google Gemini AI for accessibility and patient understanding.
Log in or sign up for Devpost to join the conversation.