Work: 2022–2024 · Company project
In microbial testing, labs need to find and count colonies, then pick individual colonies for further analysis. Growing the colonies and checking them takes time and manual work.
Microsensor Labs developed MagiCyte MB to automate these steps. The system handles samples, incubates them, takes images over time and picks colonies. Image analysis tracks their growth and helps decide which colonies to pick.
My work on colony analysis
As Co-founder and Director of Algorithm Development at Microsensor Labs, I worked on the algorithms that find, count and select microbial colonies in these images.
- Aligned images taken at different times to account for changes in camera position.
- Used computer vision and machine learning to detect possible colonies and track their growth.
- Developed algorithms to select individual colonies for automated picking and further analysis.
This is a proprietary company project.