Joint Particle Detection and Analysis by a CNN and Adaptive Norm Minimization Approach

Area of Science:

• Biomedical Engineering
• Signal Processing
• Nanotechnology

Background:

• Optical flow cytometry is complex and costly.
• Magnetic flow cytometry offers a simpler alternative using magnetic nanoparticles.
• Accurate cell counting and parameter analysis are crucial for flow cytometry.

Purpose of the Study:

• To develop a novel signal processing framework for magnetic flow cytometry.
• To enable joint detection and analysis of single and multi-cell signals.
• To improve the accuracy and efficiency of cell concentration and parameter determination.

Main Methods:

• Utilized deep learning and compressive sensing techniques.
• Developed a framework for processing both single-cell and superimposed signals.
• Employed adaptive norm minimization for signal analysis.

Main Results:

• Achieved cell counting with less than 2% relative error.
• Demonstrated reliable inference of cell parameters even at high noise levels.
• Validated the framework on simulated and experimental data from polymer microparticles.

Conclusions:

• The proposed framework enhances magnetic flow cytometry capabilities.
• It offers a robust solution for analyzing complex biological samples.
• This approach paves the way for more accessible and accurate cell diagnostics.