SymptoChecker

Thumbnail

Inspiration

What it does

Inspiration

Since the beginning of this pandemic, the world is facing is a shortage of proper medical aid for patients.

What it does

So we have created a solution using which we can improve the convenience for both patients and doctors by enabling remote communication.

How we built it

The app uses Infermedica API that asks the user to enter his/her symptoms, and then we ask some questions according to the signs that patient reports. The general questions include patient traits like BMI, behavior, geolocation(to identify location-specific diseases), etc. Then some specific questions are asked, which are based on previous responses. After answering all the answers, a set of possible diseases is displayed with its corresponding probability. A 5 level variable 'triage' is given to the session depending upon the diseases which signify the requirements of treatment for patient (ranging from self-care to emergency-ambulance). This value is sent to the admin side of the app (which is accessible by hospital authority) to assist the patient accordingly. Note that the hospital gets the name of the most probable disease, and hence doctors of the concerned field can be informed about the patient's conditions beforehand. In the case of usual cases (like common-cold), the app also makes "telecommunication_applicable" as valid, making sure that patients don't need to arrive at the hospital for a consultation, which further implies that doctor can engage in more severe cases which needs attention.

Challenges we ran into

On the way to create this project, we had to go through Infermedica API to use it to full potential. We also had to make API where the user side of the app sends a post request to keep a log of all the problems that are being reported. It took us a while to determine how the API should work and how it should appear to the hospital management side. However, we overcame this problem by using MongoDB, which made things easier for us. We later connected it to the hospital side of the app using get requests (It shows all the relevant information). Another problem that we are facing right now is using our COVID model to deploy an app on Heroku.

Accomplishments that we're proud of

In the long run, it can help create a history of the patients, which would be helpful for treatments of patients hence improving patient-doctor communication. We succeeded in our objectives which were as follows:

We ended up completing our prototype and explain our idea effectively (We guess we did :P).
To make doctor-patient communication more effective
To keep track of the history of the patient
Prioritizing the cases according to the seriousness
Helping the users understand the diseases in a more effective way

What's next for SymptoChecker App

Turning this web app into a mobile application
Implementing our COVID ML model for chest X-Ray
Adding an option to register and OTP verification for user login
Making it available as a chatbot on popular chatting platforms like telegram and messenger
Integrate the app to take data like pulse rate, step-count, sleep routine, etc. from smartwatches and make it an all-round fitness companion