Introduction

This project analyzes Computed Tomography lung studies exported as .mp4 videos using OpenCV in order to return the following results:

• Total pulmonar damage
• Overall damage score (see)
• CO-RADS score (see)

This is accomplished by using a series of image processing operations over each frame of said study, among the most important tasks are:

• Isolation of pulmonar area

• Detection of tissue

The damage is approximated using a countNonZero() operation over the detected tissue map, and adjunsting for the average % of tissue that should be inside the lungs; this approach is fundamentally flawed as damage will be detected even in a healthy individual, as such, an overall damage of <15% could be considered as being effectively 0%.

Usage

Setup

git clone https://github.com/catneep/covyx
cd covyx
python3 -m venv venv    # python for Windows
. venv/bin/activate     # venv/scripts/activate for Windows
pip install -r requirements.txt

Or by using make:

git clone https://github.com/catneep/covyx
cd covyx
make

Analysis

Syntax:

python3 . {path} --time --pretty

Where:

• path: The path to the .mp4 file to analyze
• --time: Add the analysis runtime to the results
• --pretty: Pretty print the results

Example:

python3 . /path/to/file.mp4 --time

The output is generated as a JSON string, that may be saved to a file by using:

python3 . /path/to/file.mp4 --time > /path/to/results.json