Quantitative 3D analysis of brain vasculature is a fundamental problem with important applications, for which vessel segmentation is a first step. Traditional segmentation methods based on parametric models have limited accuracy. More recent techniques based on machine learning have promising results but limited generalization capability. We present a deep-learning based segmentation method that overcomes limitations of existing systems and demonstrates the ability to generalize to various imaging setups, samples including both in-vivo/ex-vivo data, with state-of-the-art results. We achieve so by exploiting several novel methods in deep learning, such as semi-supervised learning. We believe that our work will be another step forward towards improved large-scale neurovascular analysis.
In this study, an automated serial two-photon microscope was used to image a fluorescent gelatin filled rodent’s brain in 3D. A method to compute vascular density using automatic segmentation was combined with coregistration techniques to build group-level vasculature metrics. By studying the medial prefrontal cortex and the hippocampal formation of 3 age groups (2, 4.5 and 8 months old), we compared vascular density for both WT and an Alzheimer model transgenic brain (APP/PS1). We observe a loss of vascular density caused by the ageing process and we propose further analysis to confirm our results.