Fixed Vs Adaptable Point Spread Function (psf) in Deconvolution Two-photon Microscopy of Pathological Neo-vasculature

Pathological neo-vascularization, i.e., the expansion of the vascular bed in solid tumors and after hindlimb ischemia and myocardial infarction, occurs through angiogenesis and/or arteriogenesis [1]. Quantification of microvascular morphology would allow evaluation of the effect of proand anti-angiogenic therapy. However, due to the point spread function (PSF) of the TPLSM setup, quantification of volumetric data is highly overestimated. Recently, we developed a method (Figure 1) to accurately quantify microvascular density, length and diameter and degree of branching of 3D tumor microvasculature data that was deconvolved blindly using a fixed PSF [2]. However, with TPLSM depths can be achieved up to several hundreds of micrometers in non-homogeneous biological tissue. The associated increased scattering of fluorescence photons with increasing depth widens the PSF and decreases image quality. Currently, we are adapting and applying a non-blind deconvolution method with a spatially dependent PSF to evaluate its performance and are expanding our quantification method to asses microvascular orientation. Figure 1: Tumor microvasculature after A) blind deconvolution, B) spatial filtering, C) isosurface application and D) conversion to a skeleton [2].