Image Analysis

From cell imaging to high-throughput counting and tracking

Manual cell counting comparison.
Lines are countings performed by different lab members.

Manual identification, counting and tracking of cells, while useful as a quick tool for smaller projects, quickly become difficult or outright impossible as the backlog of images increase to the hundreds or thousands. This process is also prone to errors and fluctuations, even when performed by trained individuals (see graph). Automation is important.

Many open source solutions such as Icy and CellProfiler exist. These software include a vast array of modules, plugins and pipelines to detect cells and analyze images.

Example of object detection as performed by CellProfiler.

    Our current work focuses on these aspects of image analysis:

  • Incorporation of programming languages into biological sciences, allowing for faster data analysis;

  • Tracking cell across multiple images, allowing reconstruction of cell destiny and colony family trees;

  • Overcoming challenges in cell segmentation software: correctly detecting cells close to each other and cell colonies;

  • Creation of an integrated & automated pipeline, from long-time image acquisition all the way to data visualization.

  • If you would like to propose a collaboration, please contact us.