Versatile Image-Assisted Cell Sorting by Selective Trapping with Spatiotemporal Multiparameter Targeting.

Area of Science:

• Biotechnology
• Cell Biology
• Microscopy

Background:

• Current cell sorting techniques are limited by versatility, complexity, cell number requirements, and object size constraints.
• Existing image-assisted sorters struggle with motion blur in high-resolution imaging.

Purpose of the Study:

• To introduce a novel cell sorting method, 2D-SIGMAT, addressing limitations of current cell isolation technologies.
• To enhance cell sorting precision, efficiency, and versatility across a range of object sizes.

Main Methods:

• Developed two-dimensional sorting with image-guided multiparameter adjustable targeting (2D-SIGMAT) using dynamic in situ light-activated cell trapping.
• Integrated high-resolution imaging capabilities, surpassing existing methods in pixel density and reducing motion blur.
• Utilized deep neural network models, including YOLOv5, for robust target detection and sorting.
• Demonstrated compatibility with fluorescent, bright-field imaging, and temporal data analysis.

Main Results:

• Achieved precise and efficient isolation of objects ranging from single cells to organoids.
• Demonstrated high-resolution imaging with over ten times more pixels per image compared to other image-assisted sorters, without motion blur.
• Showcased sorting based on high-resolution temporal data, capturing dynamic cellular behavior.
• Attained up to 98% recovery efficiency at a throughput of 2000 cells per second using YOLOv5.

Conclusions:

• 2D-SIGMAT transforms standard microscopes into versatile, high-performance cell sorters with scan-select capabilities.
• The method offers broad application potential in various biological research fields.
• 2D-SIGMAT overcomes major limitations of conventional cell sorting, enabling advanced cellular analysis.