“…To test the accuracy of the segmentation pipelines, the L -nets were then used to simulate in vivo photoacoustic vascular networks embedded in muscle tissue using the Simulation and Image Processing for Photoacoustic Imaging (SIMPA) python package (SIMPA v0.1.1, https://github.com/CAMI-DKFZ/simpa ) [47] and the k-Wave MATLAB toolbox (k-Wave v1.3, MATLAB v2020b, MathWorks, Natick, MA, USA) [48] . Planar illumination of the L -nets on the XY plane was achieved using Monte-Carlo eXtreme (MCX v2020, 1.8) simulation on the L -net computational grid of size 10.24 × 10.24 × 2.80 mm 3 with 20 µm isotropic resolution.…”