bessouat40 / nginx-reactjs-fastapi-pneumonia-detection

Pneumonia Detection Software with Database

Webapp : Pneumonia Detection

This is a software to detect pneumonia from a chest X-ray. You can treat several patients at the same time and then you can store results in a PostGreSQL database. You have access to this database and you can search someone with is name.

USAGE

This software is easy to use.

First clone the repo :

git clone https://github.com/Bessouat40/nginx-reactjs-fastapi-pneumonia-detection.git

Then rename env file :

cp .env.example .env

And now you can run the app :

• First, up database : docker-compose build db && docker-compose up db and wait until database is ready

• Then up entire software : make start

You can see your PneumonIA running at http://localhost:3000/.

AI Model

It's a fine tunned VGG16. https://github.com/Bessouat40/unsupervised-image-anomaly-detection

nginx-reactjs-fastapi-pneumonia-detection's Issues

Add dropzone area

Correction before storing data

Give the doctor the possibility to overwrite IA prediction (void mistakes)

Change prediction values in csv

Change 0 to normal and 1 to pneumonia

Modify prediction form in csv

Readme modification

Modify link to GitHub repository

Zoom when clicking on an image

Add AI explaination

Add injection date in database

Use .env

Add elastic database (or SQL)

Save : patient name + diagnostic

Add sidebar

Add keacloak

add help section

Add use cases for a better understanding of the app

Table modification

Add multiple pages (with for example 10 lines on each page)

Change AI model

Use cnn or something else

Add dashboard

Number of patient in database, part of pneumonia/normal,...

Add injections in elastic database

2 services : cluster elastic + app
Possibility of database injection in pre existing elastic

Add database management

Possibility to delete informations/add date information when storing datas

Add doctor name to database

Create themes

Add gitlab ci

Add makefile

Rajouter cadre autour des composants + fond de couleur pour l'app

Add unittest

Add multiple filters

Rajouter pastilles

Ajouter une pastille verte ou rouge sur la description des images (à côté du nom de l'image)