sagban / pneumonia-ppe

PneumoScan: An AI Tool For PPE Check & COVID19 Testing

INTRODUCTION

PneumoScan.ai is developed to be a secured AI tool with the purpose to assist radiologists with COVID-19 dectection on chest X-ray, and medical staff's PPE safety check amid this COVID-19 pandemics.

INSTALL

Now Let's Install Virtualenv with pip. There are other ways also to install virtualenv but I prefer this.*

$ easy_install pip

Next step is to install the virtualenv package:

$ pip install virtualenv

Great you have installed virtualenvon your machine.

Cloning the project. Type this command in terminal to clone the repo.

$ git clone https://github.com/sagban/pneumonia-ppe.git

Create an Environment with virtualenv

$ cd pneumoscan-ppe

To create the environment with virtualenv:

$ virtualenv python venv  #see alternative if you are using other than LINUX/UNIX.

After creating virtual environment, it's time to activate it. Type this command

$ source venv/bin/activate

To check wether the cloning process done corectly type ls,and it'll look like this.

ls
app  covid-19-model  db.sqlite3     manage.py   pneomonia               ppe-check   requirements.txt    venv

$ pip install -r requirements.txt

Now, start the deployment server

$ python manage.py runserver

If everything worked fine >> Congratulations, you setup the pneumoscan project in your pc.

Description

Here, is the description of the folders containing files used for the deployemnt of the pre-trained model (covid-19 model), VitechLab’s PPE Detector for Laboratory Safety (curated model) and Quantipil’s Barcode/QR-code Scanner (curated model) from AWS Marketplace.

1. covid-19-model/ This folder containes the files related to the pre-trained tensorflow model deployed on the aws sagemaker. We used the tesnsorflow model checkpoints o deploy a TensorFlow model using Amazon SageMaker, taking advantage of Amazon SageMaker deployment capabilities, such as selecting the type and number of instances, performing A/B testing, and Auto Scaling. There are two main files here, covid19-lambda-function.py covid-endpoint.ipynb
   covid-endpoint.ipynb: It is used for the creating the endpoints from the Tensorflow saved checkpoints and them converting to the Tensorflow Protobuff. After that model is move to the S3 bucket for further deployment of endpoints using the sagemaker.
   covid19-lambda-function: This file containes the code used for the setting up the aws lambda function to invoke the created endpoints.
   Here is the model.tar.gz saved in the S3 bucket: https://sagemaker-us-west-2-453101909370.s3-us-west-2.amazonaws.com/model/model.tar.gz

2. ppe-check This folder containes the files related to the VitechLab’s PPE Detector for Laboratory Safety (curated model) from the marketplace.

3. badge-scan This folder containes the files related to the Quantipil’s Barcode/QR-code Scanner (curated model) from the marketplace.

Happy Coding!