Coherent anti-Stokes Raman scattering cell imaging and segmentation with unsupervised data analysis

Area of Science:

• * Biophotonics and advanced microscopy techniques.
• * Label-free cellular imaging and analysis.
• * Multivariate data analysis in biological systems.

Background:

• * Coherent Raman imaging is a powerful tool for live-cell analysis, probing molecular content with high speed and sensitivity.
• * Multiplex coherent anti-Stokes Raman scattering (MCARS) offers label-free bioimaging with high spectral resolution and reduced acquisition times.
• * Current methods often require complex data processing or phase retrieval.

Purpose of the Study:

• * To develop a novel, unsupervised methodology for hyperspectral cell imaging and segmentation.
• * To integrate MCARS imaging with multivariate curve resolution (MCR) using spatial constraints.
• * To demonstrate a simplified workflow for label-free cellular analysis without phase retrieval.

Main Methods:

• * Multiplex coherent anti-Stokes Raman scattering (MCARS) microspectroscopy with sub-nanosecond laser pulses.
• * Unsupervised data analysis using multivariate curve resolution (MCR).
• * Implementation of MCR with signal non-negativity and novel cell segmentation-based spatial constraints.

Main Results:

• * The MCARS-MCR approach successfully segmented and analyzed various cell types (HEK293, C2C12) and physiological states (interphase vs. mitotic, live vs. fixed).
• * Demonstrated sensitivity in detecting molecular changes in cells overexpressing TrkB with and without BDNF.
• * Validated segmentation capabilities for single cells and cell clusters.

Conclusions:

• * The proposed MCARS-MCR methodology provides a robust, unsupervised, and label-free approach for hyperspectral cell imaging and segmentation.
• * This technique simplifies complex bioimaging analysis, offering broad applicability in cell biology and disease research.
• * The integration of spatial constraints in MCR enhances segmentation accuracy and analytical efficiency.