SYNAPSE

Inspiration

Modern radiologists face growing workloads and fragmented workflows—constantly switching between PACS viewers, AI dashboards, and annotation tools. Synapse was created to fix that. Our team set out to build an AI-powered collaboration platform that seamlessly connects radiologists and machine intelligence in one unified workspace. Overreliance on AI is also another prevalent issue faced in the medical industry; we wanted to ensure that the radiologist's insight comes first.

What it does

SYNAPSE provides a unified workspace where doctors can: upload and view DICOM images directly in the browser, annotate scans with intuitive tools for highlighting and labeling findings, run AI analysis using Hoppr’s chest X-ray models, which detect abnormalities and generate interpretable insights in real time, compare human and AI results side-by-side, or merge them to find consensus regions.

How we built it

We built the platform using React 18 + TypeScript + Vite on the frontend and Python 3 + Hoppr SDK on the backend.

Challenges we ran into

Along the way, we tackled challenges like DICOM rendering and localization of the AI diagnosis.

Accomplishments that we're proud of

We’re proud of how Synapse evolved from a concept into a fully functional prototype in such a short time. Despite working across multiple languages and frameworks, we built a complete AI-assisted radiology workflow that actually works — from DICOM upload to AI inference to collaborative annotation.

What we learned

We learned how to integrate the HOPPR API, learning Flask, REACT, Tailwind, and much more along the way.

What's next for SYNAPSE

Next steps for SYNAPSE include fine-tuning AI models using real radiologist feedback to improve accuracy and adaptability across diverse patient populations and expanding modality support to include CT, MRI, and ultrasound imaging.