CANCOL, a Computer-Assisted Annotation Tool to Facilitate Colocalization and Tracking of Immune Cells in Intravital...

Area of Science:

• Immunology
• Biomedical Imaging
• Computational Biology

Background:

• Two-photon intravital microscopy (2P-IVM) is crucial for studying cell interactions in vivo.
• Automated cell tracking in 2P-IVM data is hindered by technical artifacts like brightness shifts and crosstalking.
• Existing machine learning methods lack specificity for intravital immune cell imaging challenges.

Purpose of the Study:

• To develop a machine learning tool, CANCOL, to improve automated immune cell tracking in 2P-IVM.
• To address limitations in cell detection and tracking caused by imaging artifacts.
• To enhance the accuracy and efficiency of analyzing complex intravital imaging data.

Main Methods:

• Developed CANCOL, a machine learning-based tool for automated immune cell tracking.
• Incorporated guided annotation to handle problematic objects in 2P-IVM data.
• Created a virtual colocalization channel specific to the cells of interest.

Main Results:

• CANCOL significantly improved the accuracy of automated cell tracking in challenging 2P-IVM videos.
• The tool reduced the time required for manual track curation.
• Validated performance on 2P-IVM data from murine organs.

Conclusions:

• CANCOL offers a robust solution for machine learning-based immune cell tracking in 2P-IVM.
• The tool enhances the reliability and efficiency of intravital imaging analysis.
• Facilitates more accurate studies of cell-to-cell interactions in living organisms.