BigDocs
This is a full-fledged Telemedicine and Healthcare Management Platform that connects patients and doctors through virtual consultations, appointment booking, medical report uploads,etc
Created on 9th February 2025
•
BigDocs
This is a full-fledged Telemedicine and Healthcare Management Platform that connects patients and doctors through virtual consultations, appointment booking, medical report uploads,etc
The problem BigDocs solves
Managing doctor appointments, virtual consultations, and prescriptions can be challenging for patients. Tracking approved appointments, accessing telemedicine links, and retrieving prescriptions often require multiple steps, leading to confusion and inefficiency.
🔥 Features & Benefits
✅ View Approved Appointments – Patients can instantly see their approved appointments, including doctor details, date, and time.
🔗 Easy Telemedicine Access – No need to search emails; patients can join their virtual consultation with a single click via /telemedicine/{roomId}.
💬 Doctor Messaging – Enables seamless communication with doctors, ensuring timely updates and better interaction.
📅 Real-Time Updates – The system fetches only approved appointments, so patients always see up-to-date information.
📝 Prescription Management – Doctors can add prescriptions, which are directly available on the patient’s dashboard, ensuring easy access to medications and treatment plans.
⏳ Improved Efficiency – Doctors can manage their appointments and prescriptions more effectively, reducing confusion and missed sessions.
🧠 AI-Powered Disease Prediction (BERT Model) – Patients can input symptoms, and our custom-trained BERT model predicts potential diseases with high accuracy. This assists both patients and doctors in early diagnosis and better decision-making.
💡 Since this was a 48-hour hackathon, we deployed a lower-parameter BERT model for quick inference. However, we are currently training a more robust and accurate version for improved disease prediction.
🚀 This feature simplifies healthcare by making appointments, telemedicine, prescriptions, communication, and disease prediction faster, more accessible, and hassle-free!
Challenges we ran into
During the development of this project, we encountered several hurdles that required innovative solutions:
🚀 Fetching Approved Appointments in Real-Time
Initially, even after approval, patient dashboards showed "No approved appointments" due to Firebase authentication delays and incorrect queries. We resolved this by ensuring auth.currentUser was initialized before fetching data and implementing better state management for real-time updates.
🔗 Generating & Displaying Telemedicine Links
We needed to create dynamic meeting links (/telemedicine/{roomId}) while restricting access to only the consulted doctor and patient. Some roomIds were not stored correctly, causing access issues. Fixing this required verifying that roomId was assigned upon appointment approval and fetching it correctly in the MessageDoctor component.
📄 Prescription Handling
Prescriptions added by doctors were not updating in real-time due to Firestore's default caching. We implemented Firestore listeners (onSnapshot) to ensure instant updates whenever a prescription was added.
⚡ Fixing Stuck Loading States
Some pages got stuck on "Loading..." due to Firebase permission issues and missing error handling. We fixed this by adding proper try-catch blocks, improving loading state management, and refining Firestore security rules to ensure proper data access.
🤖 Deploying the ML Model on Hugging Face with Flask & Docker
Our BERT-based disease prediction model worked locally but faced issues in deployment due to dependency mismatches and memory constraints. We resolved this by:
✔️ Using Flask to serve the model via an optimized API.
✔️ Containerizing the solution with Docker for better portability.
✔️ Deploying on Hugging Face Spaces with a lightweight execution strategy.
By tackling these challenges, we built a robust, AI-powered healthcare system that ensures real-time efficiency and accessibility. 🚀