Mobile application for Healthcare using Machine Learning

Our client works with government entities for managing health data of the rural areas, especially the children from government schools. This helps government to detect any major diseases in its initial level and provide proper neutrition to the children.

Client's requirement

They wanted to develop an app for a least trained person who will go to all these schools and collect the health data. The android app should detect various body parts and such as eyes, tongue, teeth, skin and nails and click the highest quality image and upload the same in their Electronic health record system.


We analyzed the requirement and found that we can detect the body parts with the help of tensorflow lite. For the image quality check such as blur image or image with low light we used opencv.

What we did?

  • We designed the UI/UX for the app
  • With the simple, minimal yet attractive UI, We fulfilled clients purpose of building an app which can be used with a person who is not trained professional or healthcare worker.
  • We integrated the tensorflow lite for the object detection, we created different different models for each object and trained them with the data we had.
  • Once we achieved that we created an overlay image which showed the shape which needed to be clicked by the user.
  • Once the image is clicked we checked its quality with the help of opencv and stored it into the local database
  • Once the image is stored we integrated the upload feature in the background where we integrated the APIs provided by the client and pushed the data to their server
  • Android mobile app

  • Login with the credentials
  • Get list of the schools which needs to be visited
  • Get student health ID and list of the students from a particular school
  • Per student we managed to click images for body parts such as eyes, tongue, nails, teeth, skin
  • with an overlay image and a guided image for the same, user can know what to do and how to click a picture.
  • Once the image is clicked it will be checked for its quality and will be stored to the local database along with that it will be pushed to the EHR
  • Technology we used

    • Android - Java
    • Services used - TensorFlow Lite

    Project Info

    client: Food delivery application like V-thru

    Budget: USD 6000 - USD 10000

    Category: End to end product development

    Date: 05 March, 2021

    Do you want to build a web app for your b2b services? Let us help you.