Easy Doc

Text based medical services for poor people who can not afford smartphones or internet service.

Comment

Inspiration

Healthcare is a huge problem for low income population in San Francisco, and medical plans are too costly to afford.

Our idea is to use text messaging to educate and help the poor get the necessary first aid information instantly.

We all have started taking internet connectivity for granted because of the fact that we can afford it. On an average, it takes $50.00 per month for about 100 Mbps of speed. But for people with unstable sources of income, sparing that much of money just for internet is not possible.

So as a community, do we only help the privileged? Do we only educate the ones who can afford an internet connection? NO!!!!!

Our solution does not depend on the internet connectivity!

Numbers clearly suggest that getting a phone is much more affordable than buying an internet plan. We did some research and were able to spot some phones to buy as cheap as $10 on eBay or craigslist. Also, donation homes receive a lot of outdated phones which can be used.

What it does

One can send a text message of symptoms or image of injury to +18646591878. As a result, our application detects based on natural language processing and image processing of what the issue might be. After the detection, it returns the step by step guide on how to do first aid. It can also detect senders location and send information about free clinics around the location.

How I built it

  1. Flask and Python: We used python Flask to build the backend which we hosted on ngrok.
  2. Twilio: Twilio receive and send a text back to sender
  3. Microsoft LUIS api: As the sender can describe their symptoms in their own words, we used Microsoft LUIS api to perform natural language processing to detect the disease based on the text.
  4. Microsoft custom vision api: We used Microsoft custom vision API to do image processing and detect the injury based on senders image in real time.
  5. Google Maps api: We used gmap api to detect the location of text message sender to return nearby free clinics.

Challenges I ran into

Coding for Twilio to receive input texts was difficult as it needed the application to be publically hosted

Accomplishments that I'm proud of

Building the project end to end! Three of the teammates were students of Hack bright academy and came from a non-CS background.

What I learned

The learning curve was HUGE! We learned how to host a flask application which took most of the time.

What's next for spuds-hackathon

We want to support texts in any language so that we are more inclusive.

Built With

Share this project:

Updates