Datascientest project - Analysis of Covid-19 chest x-rays

Project description

In order to diagnose patients with Covid-19, the analysis of chest X-rays is a possibility to be explored to more easily detect positive cases. If the classification through deep learning of such data proves effective in detecting positive cases, then this method can be used in hospitals and clinics when traditional testing cannot be done.

Resources to refer to:

Data:

The data set contains chest x-ray images for covid-19 positive cases but also x-ray images of normal and viral pneumonia. Link to dataset: https://www.kaggle.com/tawsifurrahman/covid19-radiography-database (Data size: 1.15 Gb)

Bibliography:

https://arxiv.org/abs/2003.13865 https://doi.org/10.1016/j.compbiomed.2021.105002

Running the project

GIT

Git commands to use in terminal/console

show git log

git log --oneline --graph --name-status

show changes by user (blame)

git blame filename

show change between commits

git diff commit1 commit2

reset change to last commit

git reset --soft HEAD~1

create new branch

git checkout -b new-branch-name

add stages files

git add file1

commit changes with a message

git commit -m "message"

Python virtual environment

To setup venv for this project, use the script venv.sh in the root folder of this project as the following:

source venv.sh help

Install project dependencies

To install project's dependencies, use Makefile in the root folder of this project as the following:

make install

Jupyter notebook kernel to the .ds_covid19 env

To set Jupyter's notebook kernel to .ds_covid19 env, execute this command after the env is activate as the following:

python3 -m ipykernel install  --user --name=ds_covid19

Pre-commit for python

To check that all your files follow proper standards, execute this command manually as the following:

pre-commit run --all-files

Run MLFlow

To start MLFlow, execute this command manually as the following:

make mlflow-start

To stop MLFlow, execute this command manually as the following:

make mlflow-stop

Run Streamlit

To run Streamlit, execute this command manually as the following:

streamlit run reports/streamlit_report/Home.py