LABEL-FREE BLOOD CELL ENRICHMENT AND QUANTIFICATION USING INERTIAL MICROFLUIDICS INTEGRATED TO AI-POWERED DIGITAL HOLOGRAPHIC MICROSCOPY

Analyzing leukocytes, platelets, and rare aggregates is increasingly interesting for clinical diagnostics due to their critical roles in inflammation and thrombosis.However, labeling methods or lysis protocols for flow cytometry lead to morphological changes and cell activation [1].Therefore, sample preparation-free imaging flow cytometry is preferred, but the low leukocyte count compared to RBCs (1:1000) results in low statistical power.This limitation necessitates extensive and time-consuming measurements that increase computational demands.To address these challenges, we introduce a rapid label-free microfluidic blood cell analysis workflow that combines inertial blood fractionation for leukocytes and platelets enrichment, digital holographic microscopy, and deep learning for single and aggregated blood cell imaging.