Ultrasound localization microscopy (ULM) offers non-invasive, deep-tissue imaging of the microvasculature by tracking millions of intravenously injected, clinically approved, individual microbubbles. In this work, we introduce a simulation–experiment framework, combining realistic vascular modeling with a large, open-access database of in vivo transcranial mouse ULM datasets that we recently released. We use this framework to develop novel biomarkers in small vessels such as pulsatility imaging, pulse-wave imaging, capillary transit time, and the detection of capillary stalls. This framework enabled the development of several technical improvement such as the track-and-localize approach and robust aberration correction. Using these improvements, we show the feasibility of mapping pulsatility and pulse-wave velocity in the entire brain in vessels as small as 30 microns in diameter. We also introduce the concept of single capillary reporters, i.e., single microbubbles that were tracked over thousands of frames from the arterial to the veinous side of the vascular tree and thus enabling the quantitative mapping of transit time and stalls in capillaries, which correlate with neuroinflammation in a LPS-challenge mouse model. Finally, we explore the translational potential of these methods in larger brains in vivo, paving the way for clinical imaging of the microvasculature.
Jean Provost (2025) studied this question.