Machine learning assisted multi-level differentiation of blood cells using multi-frequency electro-mechanical impedance cytometry

A multi-frequency microfluidic electro-mechanical impedance cytometer (800 kHz–8 MHz) enables machine-learning blood cell differentiation with a minimum F1 score of 0.92.

• Why it matters: Accurate blood-cell classification is needed for medical diagnosis, where both false negatives and false positives can be harmful.
• What they did: The team measured electrical and mechanical properties with multimodal impedance features, including a constriction-equipped channel and multiple microelectrodes, then trained a workflow to classify cells and rank the top 5 parameters from 70+ features.
• The result: In blood cell differentiation experiments, the classifier achieved a minimum F1 score of 0.92 and detected two populations within a mixed neutrophil and PBMC sample.

AI summary of the abstract · may contain errors · not medical advice · read the paper.