Advanced cardiotoxicity profiling using field potential imaging with UHD-CMOS-MEA in human iPSC-derived cardiomyocytes.

Accurate assessment of cardiotoxicity using human induced pluripotent stem cell (iPSC)-derived cardiomyocytes is critical for ensuring drug safety during preclinical development. However, existing in vitro methodologies predominantly focus on QT interval prolongation and arrhythmia risk, often lacking the capacity to capture the complex interplay among multiple ion channels or to detect early manifestations of chronic cardiotoxicity-both of which are essential for evaluating long-term cardiac safety. Moreover, reliable prediction of pharmacological mechanisms of action remains a significant challenge. In this study, we employed field potential imaging utilizing an ultra-high-density complementary metal-oxide-semiconductor microelectrode array (MEA) comprising 236,880 electrodes distributed across a 5.9 × 5.5 mm active area. With 91.9% surface coverage by 11 μm electrodes spaced at 0.25 μm, the platform achieves near single-cell resolution across the entire cardiomyocyte monolayer. This system enabled the extraction of high-resolution electrophysiological endpoints, including the number and spatial variability of excitation origins, conduction velocity, and propagation area-thereby extending the analytical capabilities beyond those of conventional MEAs. Pharmacological testing revealed compound-specific alterations: Isoproterenol increased excitation origins, mexiletine reduced conduction velocity, and E-4031 diminished propagation area. Although these agents are well characterized, their effects were visualized with unprecedented spatiotemporal resolution, reflecting their underlying mechanisms of action. Multivariate analysis incorporating both conventional and novel endpoints enabled accurate classification of mechanisms under acute conditions. Furthermore, chronic cardiotoxicity induced by low-dose doxorubicin (0.03 μM) was sensitively detected within 24 h-earlier and at lower concentrations than previously reported-based on significant reductions in conduction velocity and propagation area. Collectively, these findings establish a high-resolution, mechanism-aware framework for in vitro cardiotoxicity profiling, offering improved predictive accuracy by capturing multi-ion channel interactions, spatial conduction abnormalities, and early signs of chronic dysfunction.