Currently, globally, organizing and retrieving patient data is bottlenecked by the lack of connected databases, paper records, complex UX of current software, training overhead, and most importantly data overload. This costs the medical industry immensely in terms of patient care, time and money. This signals the need for an efficient, unified platform for data retrieval and storage.
What it does: PanaceaBot
What if you as a doctor or patient, could go immediately from chatting with one of your friends on facebook to a chatbot where you could ask it for any medical data you need? No need to parse through complex UX, paper records, and to waste time trying to get this data.
On an encrypted line, the PanaceaBot utilizes Natural Langauge Processing to let you text with it like you would talk to your doctor or your patient. As a patient, you could say, "Hey Panacea, what was the test the clinician ran on me? What were those results?". The bot would then message you with the answer: "Insulin level test. 8.8 mIU/ml, safe level.". You could look up any of your data that you would want.
While for the doctor, PanaceaBot is designed to be their data storage and retrieval system. Let's say a doctor wanted to check on one of his patients, but he didn't remember the what prescriptions he prescribed. Instead of having to search through their records program, software or paper, the doctor could just text PanaceaBot: "Hey PanaceaBot, what did I prescribe Peter John during our last session?". The bot would reply: "11/01/17: 1000 mg Acetaminophen every six hours for 1 week.". This would not only save hours for the patients and clinicians, this would translate into a clear profit for hospitals and facilitate adoption.
How we built it
We created a Facebook Messenger as our front-end, MongoDB for back end and created a NodeJS application with Heroku as our pipeline. We use Messenger API to display data based on the user's request through fetching objects from the Mongo Database.
Challenges we ran into
Accomplishments that we're proud of
Natural Language Processing, Messenger API
What we learned
NodeJS, Messenger API
What's next for Panacea
While we've made a lot of progress on PanaceaBot through the 24 hours, we would love to integrate more features into the PanaceaBot. Specifically, we'd like to give the bot the ability to take in speech and interpret it. In addition, we'd like the bot to have the ability to take pictures of paper medical records and to parse data from the paper into our database.