High-throughput time-stretch imaging flow cytometry for multi-class classification of phytoplankton

Area of Science:

• Optofluidics
• Biophotonics
• Single-cell analysis

Background:

• Time-stretch imaging offers high-throughput flow cytometry but faces challenges in image resolution and unraveling cellular heterogeneity.
• Accurate classification of diverse cell populations is crucial for life sciences and diagnostics.

Purpose of the Study:

• To develop an optofluidic time-stretch imaging flow cytometer for high-resolution imaging and multi-class cell classification.
• To demonstrate the utility of intracellular morphology features for improving classification accuracy.

Main Methods:

• Implemented an optofluidic time-stretch imaging flow cytometer for ultrafast, continuous operation.
• Employed comprehensive feature extraction and selection, including intracellular texture/morphology and Fourier space analysis.
• Utilized multi-class support vector machine (SVM) for classification.

Main Results:

• Achieved high-resolution time-stretch imaging, enabling visualization of sub-cellular complexity.
• Demonstrated a classification accuracy of 94.7% for up to 14 classes of phytoplankton.
• Showcased the ability to identify sub-populations using advanced feature domains.

Conclusions:

• High-resolution time-stretch imaging significantly improves single-cell classification accuracy by leveraging intracellular morphology.
• The developed optofluidic system facilitates high-throughput screening and classification of diverse cell types.
• This technology holds promise for applications in marine research, biofuel development, and biomedical diagnostics.