Imaging deep haemodynamics non-invasively remains a quest. Although optical imaging techniques can be used to measure blood flow, they are generally limited to imaging within ∼1 mm below the skin’s surface. Here we show that such optical diffusion limit can be broken through by leveraging the spatial heterogeneity of blood and its photoacoustic contrast. Specifically, successive single-shot wide-field photoacoustic images of blood vessels can be used to visualize the frame-to-frame propagation of blood and to estimate blood flow speed and direction pixel-wise. The method, which we named photoacoustic vector tomography (PAVT), allows for the quantification of haemodynamics in veins more than 5 mm deep, as we show for regions in the hands and arms of healthy volunteers. PAVT may offer advantages for the diagnosis and monitoring of vascular diseases and for the mapping of the function of the circulatory system.

非侵入性地对深层血流动力学进行成像仍然是一个探索。虽然光学成像技术可用于测量血流，但它们通常仅限于在皮肤表面以下约1 毫米范围内成像。在这里，我们表明，可以通过利用血液的空间异质性及其光声对比度来突破这种光学扩散极限。具体来说，血管的连续单次宽场光声图像可用于可视化血液的帧到帧传播，并按像素估计血流速度和方向。我们将这种方法命名为光声矢量断层扫描 (PAVT)，它可以量化深度超过 5 毫米的静脉中的血流动力学，正如我们对健康志愿者的手和手臂区域进行的展示一样。PAVT 可能为血管疾病的诊断和监测以及循环系统功能的绘制提供优势。