Cerebral blood flow recorded at high sensitivity in two dimensions using high resolution optical imaging

Knowledge about sensory-evoked blood-flow changes is essential for constraining hemodynamic response models used to interpret functional brain imaging signals, such as fMRI. Here, we extracted 2-dimensional blood-flow and its temporal modulations from high-resolution optical imaging data in the awake monkey. Optical imaging allows to track moving erythrocytes (or small clusters thereof), thus providing, albeit noisy, information about their velocity in individual blood vessels, across the whole imaged area. Here, we illustrate the algorithms that allowed us to extract, at the single microvessel level, red blood cell (RBC) motion information from the noisy optical signals. For this purpose, we developed an algorithm that is both robust and computationally efficient, using the structure tensor, known to detect an average direction of image intensity gradient. This structure tensor tool is applied to detect trajectory directions in the spatio-temporal data. Since blood-flow modulation by the cardiac pulsation was clearly detected, our method should be applicable also to study blood-flow modulations by neuronal activity, and their spatio-temporal patterns