Technology is often BY college educated people FOR college educated people. We want to extend healthcare and esp. early support for Covid-19 to as many people as fast as possible.
I'm from Tier-3 India and have seen growing up how little health care people have there with less than 1 doctor in 10 km radius. This model will help those in most need.
What it does
Anyone in over 10 most spoken language worldwide can click or tap 1 button and give voice feedback of their health condition through their €15 Android's mobile browser. No app, no chatbot, nothing. Health Brik Machine Learning detects Covid-19 triage keywords ("I have dry cough") and surfaces alerts to Doctors at medical clinics who can prioritise patients for next steps like hardening quarantine, sending for testing or stopping immuno-suppresant medication etc. Early intervention can break chain transmission.
Important: This is a human centric doctor led model. Your recommendation comes from a real doctor and not a chatbot.
How I built it
We (team of 5) built the main platform using AWS services and mocks in Adobe. Front end is angular.js, data table etc at the backend will be built during the hackathon.
Challenges I ran into
Finding doctors to create the mockup test the prototype. (we have a few, looking for more). Finding medical-legal experts for data privacy compliance (found after weeks of trying)
Accomplishments that I'm proud of
We are building a fast implementable but watertight model that can be scaled quickly for those who don't get benefits of technology the most
- Symptom recognition scalability in 10 languages figured out (build in English first but path same for all other languages)
- Human centric doctor led model and hence doctor friendly mockups done with their feedback
- Spoke with 2 national health care agencies who would be willing to recommend this to doctors in medical clinics
- 100% Compliance to data safety norms - Got checklist ready for Medical GDPR, US HIPAA/Hitech/Safe Harbor and Indian Telemedicine act ready.
What I learned
- Be more utility driven to scale and include people: Our model has no fancy wearable, chatbots nothing
- Test, learn from mistakes and move fast: 'Perfection' is the enemy of 'done'.
What's next for Health Brik
- Finding a small investment to recruit a few key resources 1. Data scientist for scalable training set 2. Medical lawyer consulting.
- More medical clinic awareness to start testing the pilot and stabilise it for a fast product release in many languages. Anthony Faucci says Covid-19 is a cyclic disease that will return in fall, which means we need to improve "social preparedness" mechanisms fast.