Fig. 7

Microvascular networks (MVN). MVN1 and MVN2 are networks extracted from mouse cortex data [19]. The vessel segment visualizations show MVN2. Pial vessels have been removed. Classification into arterial and venous vessels is based on pressure maps computed with a vascular graph model (see Methods and Fig. 9). (A) Vessels with \(r_{\textrm{v}} \ge 3\,\upmu \hbox {m}\) are shown in red (arterial) and blue (venous); smaller vessels are translucent gray. Segments are rendered as cylinders with radius \(r_{\textrm{v}}\). (B) The color map is scaled by the estimated filtration coefficients \(L_p\) shown for each vessel segment. Segments are rendered as cylinders with radius \(r_{\textrm{o}}\). (C) For each of the 6 analysis layers (layer 0 is closest to the pial surface) of \(200\,\upmu \hbox {m}\) vertical thickness (\(100\,\upmu \hbox {m}\) for layer 5) and both networks, the filtration capacity, the average \(L_p\), the diffusion capacity, and the average \(C_M\) for each vessel category (A: arterial, V: venous, C: capillary vessels) is shown. The surface area of each vessel segment has been computed by assuming a cylinder with radius \(r_{\textrm{o}}\). The averaged segment lumen radius (\(r_{\textrm{v}}\) from [28]) and the averaged estimated outer endfoot sheath radius (\(r_{\textrm{o}}\)) over depth are shown in the rightmost figure. Missing data points correspond to \(\sum S=0\)