Skip to content

An open-source project for applying deep learning to medical scenarios

License

Notifications You must be signed in to change notification settings

smaranjitghose/auto_vaidya

Repository files navigation

Auto Vaidya

An open source solution for creating end-end web app for employing the power of deep learning in various clinical scenarios like implant detection, pneumonia detection, brain mri segmentation etc.

Suggestions for PR:

  • Please give your PR for the test branch unless requested otherwise by the project maintainer
  • Name your PR appropiately
  • Ensure that you had already raised an issue for this PR and the project maintainer had approved and assigned you
  • In the PR description, typically the following are expected:
    • Dataset Used:
    • Dataset Size:
    • Dataset Source:
    • Link to Colab Notebook: Please ensure you give access for view to anyone with link
    • Your Exploratory Data Analysis [Snapshots of the relevant ones and your inference from that]
    • Any Pre-Processing methods used. [Elaborate on them]
    • Your framework to train
    • Different methods used for training
    • Test/Train Split
    • Results: Please do not simply state test accuracy. Other perfomance metrics like F1 score,etc are expected
    • ** Draw a table to show the comparitive analysis of the performance of the different methods you used
    • Conclusion: Which method you think is best and why?
  • A copy of the notebook used for your training is expected inside the notebooks/ directory.
  • Please name the notebook as name_of_the_problem_your_github_username
  • The model files are expected to be inside a models\name_of_your_problem\ directory
  • If you are using TensorFlow 2.0, please give both the h5 as well as saved_model file
  • Once your PR, gets approved uptil this, proceed with a follow up pr to integrate it inside the streamlit app. Refer this if you are unaware of how to use streamlit and host it
  • For the streamlit app, it would be a good practice if you define the function for classification/prediction/regression inside a separate python file say your_problem_name.py and import it inside app.py ( Believe me this would save a lot of time otherwise wasted in debugging)
  • For the second PR, you are expected to do the above changes and provide screenshots/a small clip of the working model of the app after integrating your model from the previous PR
  • For the second PR, it should be one the test branch only, later the project maintainers will merge it with the master branch for a stable release
  • For PRs, related to frontend please give it to the frontend branch
  • Once accepted, give a follow up PR to the test branch to render your html,css files for a page using streamlit
  • As stated above you are expected to give screenshots, descriptions and other details for the PR

Entire App on Heroku: https://auto-vaidya.herokuapp.com/ Frontend on Netlify: autovaidya.netlify.app