-
-
Automated medical report generation with downloadable PDF and AI-recommended appointment slot selection.
-
Personal health dashboard showing reports, metrics, and trends with AI-driven analysis of uploaded medical data.
-
AI-assisted appointment booking system with smart scheduling based on predicted disease and urgency level.
-
Multimodal consultation using voice input and medical image upload for AI-based disease prediction and guidance.
-
Secure user authentication system with login/signup for personalized AI-powered healthcare assistance.
-
Automated email confirmation with attached AI-generated medical report using SendGrid integration.
Inspiration
Access to quick and reliable healthcare support is still limited, especially when early diagnosis can make a huge difference. We wanted to build an AI-powered medical assistant that feels like a real doctor—capable of understanding symptoms, analyzing medical data, and guiding patients all the way to booking an appointment.
What it does
The AI Medical Agent for Diagnosis & Appointment Booking is a multimodal healthcare assistant that:
Accepts text, voice, and medical images (X-ray, ultrasound) in multiple languages
Uses AI reasoning to analyze symptoms and ask follow-up questions
Provides voice-based responses for accessibility
Generates a detailed AI medical report (PDF)
Key Features:
4 Deep Learning Models for disease detection + severity scoring
Healthcare Dashboard to track BP, sugar, cholesterol + previous reports
Doctor Categories: Kidney, Orthopedic, Brain Tumor, Pneumonia, General
Smart Appointment Booking with real-time slot selection
Automated Email System sending report + booking details to doctor & patient
How we built it
Integrated multimodal AI (text, voice, image) with a reasoning engine
Used deep learning models for medical image analysis
Built a backend system for report generation, scheduling, and data tracking
Implemented speech-to-text and text-to-speech for interaction and in multiple languages
Automated PDF report generation + email delivery
Challenges we ran into
Handling multimodal data integration smoothly
Ensuring accurate medical predictions with limited datasets
Designing real-time appointment flow with AI-generated outputs
Maintaining a balance between speed, accuracy, and usability
Accomplishments that we're proud of
Built a complete end-to-end AI healthcare system integrating diagnosis, medical imaging, report generation, and smart appointment booking into one seamless platform
Developed multimodal intelligence (text, voice, image + multilingual) with AI reasoning that understands symptoms, asks follow-up questions, and provides interactive voice responses in multiple languages
Implemented 4 deep learning models to analyze X-ray and ultrasound images, generating accurate predictions along with a severity score
Enabled real-time AI-generated medical reports (PDF) with embedded images, and a health dashboard to track BP, sugar, cholesterol, and access previous reports
Created a fully automated clinical workflow with doctor specialization, slot-based booking, and email notifications to both patient and doctor with reports attached for their reference
What we learned
Combining AI models with real-world workflows is challenging but powerful
Importance of user experience in healthcare applications
Handling multimodal AI systems requires careful design
End-to-end integration is as important as model accuracy
What's next for AI Medical Agent for Diagnosis & Appointment Booking
Improve model accuracy with larger medical datasets
Add real-time doctor consultation (video/AI-assisted)
Integrate with hospital systems & wearable devices
Enhance personalized health monitoring and predictions
Built With
- amazon-bedrock
- cloud-deployment-(digitalocean)
- computer-vision
- docker
- elevenlabs-(text-to-speech)
- faiss-vector-database
- gradio
- groq-(llama-models)
- keras
- langchain
- natural-language-processing
- openai-embeddings
- python
- rest-apis
- sendgrid
- sqlalchemy
- sqlite
- tensorflow
- whisper-(speech-recognition)
Log in or sign up for Devpost to join the conversation.