Web platform for annotating medical images and generating pathologies on studies of healthy patients.
MVP
of theFutureOfMedTech
team, developed for theLCT 2022
hackathon.
Clone
git clone https://github.com/GVatest/LCT-MVP
cd hedge-fund
Install
pip install -r requirements.txt
Start
Create database
python manage.py migrate
Start server
python manage.py runserver
Install
npm install
# or
yarn install
Start
npm run dev
# or
yarn dev
- API was developed based on
django rest framework
usingsqlite3
database, as it is an mvp project, working withsqlite3
was easier and faster than its more advanced alternatives such aspostgresql
. - Authorization is implemented based on
JWT
tokens - The frontend is implemented using the
React
library in conjunction with theRedux
state manager - Loading and processing of
Dicom
slices on the front end is done using the open-source library[DWV] (https://www.npmjs.com/package/dwv)
- The functionality of the annotator is based on the functionality of the library for working with computer vision algorithms
[OpenCV.js] (https://docs.opencv.org/4.6.0/d5/d10/tutorial_js_root.html)
I was the member of FutureOfMedTech
team at the 2022 hackathon. As a full stack
developer I took responsibility for writing the API, designing the database, and writing individual front-end modules: set up JWT
authentication, loading, primary processing and displaying Dicom
slides, participated in writing annotator functionality: annotation tools, import and export of annotation to Json
, multiple image loading, image scrolling. Also took responsibility for merge of three application modules: pathology generation module, Dicom
image annotator, studies and staff management system. The final challenge was to deploy and configure the application on a remote server, for this I used nginx
+ gunicorn
.
Contributors:
- 👤 Vasiliy Ganja
Github
: @Gvatest
- 👤 Maxim Kirilyuk
Github
: @Werserk
- 👤 Balai Micheeva
Github
: @Balaishka