Computer-aided quantification of microvascular networks: Application to alterations due to pathological angiogenesis in the hamster

•An automatic image processing pipeline to analyze intravital-microscopy fluoroscopy images of the hamster cheek pouch is presented, called PN-method. Comparisons with an intensity-threshold based method are performed.•Several image-based quantification indexes are used to characterize microvascular alterations.•The PN-method renders an accurate and robust segmentation of blood vessels for different microvascular scenarios, normal and pathological.•Images with angiogenesis can be differentiated from controls using a set of quantification indexes.•All segmentation-independent indexes (RFU, H, AvF) presented potentiality for identification of normal and neo-vascularized microvasculature.

Angiogenesis is both a physiological and a pathological process of great complexity, which is difficult to measure objectively and automatically. The hamster cheek pouch (HCP) prepared for intravital-microscopy (IVM) has been used to characterize microvascular functions in many studies and was chosen to investigate microvascular characteristics observed in normal non-infected hamsters as compared to those HCPs parasitized by Trypanosoma cruzi. Images of HCPs captured at IVM were subjected to computer based measurements of angiogenesis and histamine-induced macromolecular (FITC-dextran) leakage with an image segmentation approach that has the capacity to discriminate between fluorescence emitted by macromolecular tracers inside the vasculature and in the extravascular space. We present such an automatic segmentation methodology using known tools from image processing field that, to our knowledge, have not been tested in IVM images. We have compared this methodology with a recently published segmentation strategy based on image intensity thresholding. Our method renders an accurate and robust segmentation of blood vessels for different microvascular scenarios, normal and pathological. Application of the proposed strategy for objective and automatic measurement of angiogenesis detection was explored in detail.