PlasmaShare

Inspiration 💡💡

India have been recently hit by a second wave of COVID-19 and everything has gone back to how it was a year ago. Whenever we open Instagram we see our friends posting stories trying to find available beds, oxygen cylinders and plasma donors for their parents and relatives. They tell us how they never hear back from anyone and even if they do it often fake news.

Aim 🎯🎯

To create a plasma donation app that allow users find verified and eligible plasma donors. The app must also allow users who have recovered from Covid to register themselves as plasma donors. The app also needs to make sure that the Covid report and information provided by a user are genuine and they are eligible to be a donor. Moreover, through verification from our end we want to reduce the load on hospitals and help the people in need due to the Covid pandemic.

How we built it 💻💻

The front end of the website was developed using Flutter ,dart ,and google maps api. The backend of the website was developed using python Flask. For storing the register users's data we used Google Cloud Firestore storage. In order to register a users as plasma donor we ask the user to provide their information and upload a Covid test report. We then checked if the fields and the information provided in the report are correct using the Azure Form Recognition Machine Learning model which we implemented with some modifications and deployed it using Google Cloud (App Engine). Through are machine learning model we checked if the format of the covid test report is correct, we extracted various field from the report such as the Name of the applicant , the email address of the hospital, the testing date,and the test result in order to check if their test report is valid and if they are eligible to donate blood. Moreover, since our MVP is verification we also send a email to the hospital with the users information and their test report for one last check and then add the user to the list of verified plasma donors.

Here is the link to the deployed machine learning model : link

Note: You won't be able to upload new reports as that can only be done through the app 😅😅

Tech Stack🤖🤖

  • Flutter and Dart
  • Google Firestore and Storage
  • Google Maps Api
  • Google Cloud (App Engine)
  • Azure Form Recognition Model
  • Python Flask

Challenges we ran into 💪💪

  • As we all know machine learning is highly dependent on the data. We were only able to collect a few COVID positive test reports and thus we weren't presently able to apply more extensive features for verification.

Future Plans 📈 📈

  • Firstly we would like to gather more COVID positive reports so that we can improve the accuracy of our machine learning model and apply more features to strengthen our verification.
  • Secondly rather than sending an email to the hospitals we would like to make a sign in page. After the sign in page a normal user will just what we currently have, however if the user is a hospital or clinic they will be able to see all the people who would have requested to become a donor. After this the hospital can simply check their report and verify them, and then the user will be added to the list of verified plasma donors.
Share this project:

Updates