BandAid: Low Bandwidth Telehealth

BandAid: Low Bandwidth Telehealth

Inspiration

Millions of people worldwide live in areas with limited or unstable internet connectivity, severely restricting their access to quality healthcare. Inspired by the potential of emerging technologies for efficient 3D reconstruction, we set out to create a telehealth solution specifically designed for low-bandwidth environments, ensuring everyone has access to reliable medical consultation.

What it does

Current telehealth methods often rely heavily on continuous video calls, which consume large amounts of bandwidth and make them impractical or even impossible for patients in regions with limited internet access. BandAid enables patients in remote or low-infrastructure areas to consult healthcare professionals by sharing a few simple images of a medical issue, such as a wound or rash. Our system instantly reconstructs these images into a detailed, lightweight 3D model. Doctors view this high-quality model, interact using advanced visualization tools, and communicate with patients using optimized voice-to-text/text-to-speech methods, dramatically reducing bandwidth requirements compared to traditional video & audio calls.

How we built it

We leveraged efficient 3D reconstruction techniques to quickly convert patient-submitted photos into lightweight 3D models, significantly lowering bandwidth requirements compared to traditional video calls. The backend, built using FastAPI and Python, handles image uploads and 3D reconstruction processes. The doctor-side interface includes a dynamic HUD showcasing live bandwidth usage. The frontend was carefully optimized to run smoothly on limited hardware and low-speed connections.

Challenges we ran into

One of the main challenges was ensuring that our 3D models were generated quickly and reliably enough for real-time interaction in low-bandwidth conditions. Extensive optimization and testing were required to achieve acceptable performance and maintain visual fidelity, including using APIs optimized to run on apple silicon ensuring that high quality 3d models can be constructed in mere seconds.

Accomplishments that we're proud of

We successfully demonstrated real-time, bandwidth-efficient telehealth consultations, significantly reducing data requirements compared to existing telemedicine solutions. Our intuitive doctor interface provides actionable insights while maintaining an accessible and straightforward patient interface, dramatically improving healthcare reach.

What we learned

We gained deep insights into efficient 3D model generation, real-time audio/text streaming, and deploying low-bandwidth applications effectively. We also learned the critical importance of designing for accessibility from the outset, especially for resource-constrained environments.

What's next for BandAid

Future steps include integrating more advanced diagnostics using AI-driven image similarity and disease prediction, improving our 3D model reconstruction pipeline, and field-testing BandAid in actual remote healthcare scenarios to further enhance reliability and usability.

Built With

• fastapi
• nextjs
• python
• typescript