Abstract
The deformability of red blood cells (RBCs) is critical for microvascular circulation and is impaired in hematological disorders such as thalassemia, a prevalent public health concern in Guangdong, China. While microfluidics enable high-precision deformability assessment, current studies lack standardization in deformation metrics and rarely investigate post-deformation recovery dynamics. This study introduces an automated microfluidic platform for systematically evaluating RBC deformability in healthy and thalassemic individuals. A biomimetic chip featuring 4 µm, 8 µm, and 16 µm wide channels (7 µm in height) was designed to simulate capillary dimensions, with COMSOL CFD numerical modeling validating shear stress profiles. RBC suspensions (10(7) cells/mL in DPBS) were hydrodynamically focused through constrictions while high-speed imaging (15,000 fps) captured deformation-recovery dynamics. Custom-built algorithms with deep-learning networks automated cell tracking, contour analysis, and multi-parametric quantification. Validation confirmed significantly reduced deformability in Paraformaldehyde (PFA)-treated RBCs compared to normal controls. Narrower channels and higher flow velocities amplified shear-induced deformations, with more deformable cells exhibiting faster post-constriction shape recovery. Crucially, the platform distinguished thalassemia patient-derived RBCs from healthy samples, revealing significantly lower deformability in diseased cells, particularly in 4 µm channels. These results establish a standardized, high-throughput framework for RBC mechanical characterization, uncovering previously unreported recovery dynamics and clinically relevant differences in deformability in thalassemia. The method's diagnostic sensitivity highlights its translational potential for screening hematological disorders.