Probabilistic independent component analysis for laser speckle contrast images reveals in vivo multi - component
Nan Li1, Galit Pelled, Nitish V Thakor
1Biomedical Engineering Department, Johns Hopkins University, School of Medicine, Baltimore, MD 21205, USA. nanli@jhu.edu
Abstract:
Brain's functional response can be studied by observing the spatiotemporal dynamics of functional and structural changes in cerebral vasculature. However, very few studies explore detailed changes at the level of individual microvessels while revealing the simultaneous wide field view of microcirculation responses to functional stimulation. Here we use a high spatiotemporal resolution laser speckle contrast imaging method, in combination with probabilistic independent component analysis to reveal the changes of cerebral blood flow pattern in response to electrical forepaw stimulation in an anesthetized rat model. The proposed method is able to pick up the response of a single vessel down to approximately 20 microm diameter in a 4mm × 4mm field of view, and automatically extract response from multiple vascular components. Two main vascular components, arteriolar and capillary responses respectively, show significantly different temporal dynamics. Overall, the experimental results from five rats reveal that the specific arteriole branch proximal to the activation sites dilate prior consistently to the increase of blood flow in the capillaries with a latency time 0.91 ± 0.05s. The presented results provide novel microscopic scale evidence of the contribution of different vascular compartments in the hemodynamic response to neuronal activation.