Multiscale photoacoustic imaging of stroke in preclinical models and future directions toward clinical translation Invited

Photoacoustic imaging (PAI) combines optical absorption contrast with acoustic detection, enabling noninvasive, high-resolution spatiotemporal observation of cerebral hemodynamics, oxygenation, and vascular structural changes, thereby emerging as a powerful tool for investigating mechanisms in preclinical stroke models. This review comprehensively highlights how PAI has been employed to track pathophysiological changes and therapeutic responses in various preclinical stroke animal models, and discusses future directions toward translating these findings from preclinical research to clinical applications. PAI provides multiscale structural and functional vascular information, ranging from single-vessel resolution to whole-brain imaging. Through this capability, PAI enables real-time observation of hemodynamic and oxygen saturation (SO 2 ) changes, as well as infarct formation, during the onset of both ischemic and hemorrhagic stroke models. Moreover, in ischemic models, PAI enables the assessment of reperfusion, oxygenation recovery, inflammatory responses, angiogenesis, and collateral circulation during the reperfusion and self-healing phases. In hemorrhagic models, it facilitates monitoring of secondary injury processes such as hematoma formation and blood degradation. However, conventional animal models face inherent limitations for full clinical translation due to anatomical and metabolic differences from humans, while skull-induced optical scattering and acoustic aberration further degrade image quality and penetration depth, limiting the clinical applicability of PAI for stroke studies. In this review, we highlight future directions in three aspects: (1) technological approaches—including acoustic aberration correction, skull optical clearing, and transparent cranial window techniques—to minimize skull-induced effects and improve imaging depth and quality; (2) extending PAI beyond a hemodynamic monitoring tool toward a comprehensive stroke assessment platform by utilizing both endogenous and exogenous contrasts to visualize metabolic and neural function; (3) the potential of integrating PAI with brain organoid platforms to more precisely recapitulate human stroke pathophysiology and enable applications such as high-throughput screening.