During the COVID-19 pandemic, I collaborated with a team to develop a machine learning algorithm that utilizes computed tomography (CT) scans (in PNG format) to detect and segment lung lesions caused by the coronavirus or other infections. This model is integrated into a web-based interface, allowing users to easily upload and analyze test files for lesion identification.

Development Process

The following video provides an overview of our development process, along with a tutorial on how to use the website interface:

Video presentation: Overview of the development process.

Presentation

Below is the technical documentation for our project, detailing the machine learning model’s architecture, data preprocessing steps, and our approach to lung lesion segmentation:

Presentation PDF of developement process.

Source Code

Explore the full source code for the project on GitHub: View on GitHub.