Dense optical flow for universal blood flow analysis in particle-enhanced imaging

Particle tracking has emerged as a powerful tool for mapping blood flow dynamics using fluorescence, optoacoustic, and ultrasound imaging modalities. However, its reliance on isolating individual particles imposes demanding requirements on imaging systems, including high frame rates, high signal-to-noise ratios, and sparse tracer distributions. Here, we introduce a computationally efficient dense optical flow (DOF) framework that estimates flow direction and relative velocity directly from intensity displacement fields, eliminating the need for explicit particle tracking. Unlike tracking-based approaches, DOF scales with image resolution and operates identically on two- and three-dimensional image sequences, as well as composite views such as maximum-intensity projections. We demonstrate its versatility across fluorescence and optoacoustic particle-enhanced imaging, including full 3D cerebrovascular datasets, showing that DOF reliably resolves vascular flow patterns even under sparse-tracer conditions. This unified, modality-independent approach offers a robust tool for hemodynamic analysis across scales, from microvascular networks to whole-organ architectures.