A Bayesian Approach for Coincidence Resolution in Microfluidic Impedance Cytometry.

Area of Science:

• Biomedical Engineering
• Analytical Chemistry
• Microfluidics

Background:

• Cell counting and characterization are crucial in medicine and science.
• Coulter-type devices offer automated cell analysis but suffer from coincidence errors, limiting throughput and accuracy.
• Coincidences occur when multiple particles pass through the sensing zone simultaneously, causing undercounting and property mismeasurement.

Purpose of the Study:

• To propose and validate a novel approach for resolving particle coincidences in microfluidic impedance cytometry.
• To enhance the accuracy and throughput of cell and particle analysis systems.
• To enable precise characterization of cell/particle suspensions across a wide range of concentrations.

Main Methods:

• Utilized a microchannel with two electrical sensing zones.
• Developed a model for signals generated by coinciding particles.
• Employed Maximum a posteriori probability (MAP) estimation for parameter identification and individual particle characterization.

Main Results:

• Achieved 97% counting sensitivity and 99% positive predictive value at 2x10^6 particles/ml using synthetic data.
• Demonstrated accurate red blood cell characterization up to approximately 2500 particles/s.
• Derived a formula for expected coincidences, showing good agreement with experimental data.

Conclusions:

• The proposed Bayesian approach effectively decomposes signals from coinciding particles into individual contributions.
• This microfluidic impedance cytometry system enhances accuracy in cell/particle counting and characterization.
• The technology is suitable for applications demanding precise analysis of cell suspensions at high concentrations.