chorus-ai / privacy_scan_tool

A Privacy Scan tools for medical records

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

ChoRUS Privacy Scan Tool

Introduction

This tool helps you identify privacy concerns before sharing your data with others. It is a part of the RASx-rad and ChoRUS project for NIH.

The project introduction presentation can be found here.

Notes:

PHI data predictions are based on statistical results. We recommend having a minimum of 1,000 records for each table to ensure accuracy.

Prerequisites:

Windows 10 or Mac with Python 3.8 or 3.9 (with Git tools)

Mac OS Installation:

1. Create or choose a folder on your workstation

2. Open a terminal window and navigate to the selected folder

cd ~/my_folder
$

3. Create a Python virtual environment

python3.8 -m venv ChoRUS_env

4. Activate the virtual environment

source ChoRUS_env/bin/activate
(ChoRUS_env) $

5. Clone the package from Github

git clone https://github.com/chorus-ai/ChoRUS_Privacy_Scan.git
cd ChoRUS_Privacy_Scan

6. Install the required Python packages

python -m pip install -r requirements.txt

7. Run the Privacy Scan Tool

python main.py

Windows Installation:

1. Create or choose a folder on your workstation

2. Open a command prompt and navigate to the selected folder

cd c:\my_folder
c:\my_folder

3. Create a Python virtual environment

python3.8 -m venv ChoRUS_env

4. Activate the virtual environment

ChoRUS_env\Scripts\activate.bat
(ChoRUS_env) c:\my_folder>

5. Clone the package from Github

git clone https://github.com/chorus-ai/ChoRUS_Privacy_Scan.git
cd ChoRUS_Privacy_Scan

6. Install the required Python packages

python -m pip install -r requirements.txt

7. Run the Privacy Scan Tool

python main.py

The user instructions can be found here.

Version Log:

V_2.28.2023 Initial version for RADx-rad and ChoRUS project

V_3.6.2023 Updated models

Acknowledgemant

This project is supported by RADx-rad DCC (1U24LM013755) and ChoRUS (OT2OD032701)

About

A Privacy Scan tools for medical records


Languages

Language:Python 99.3%Language:Dockerfile 0.7%