Cerebral Vessel Extraction: From Image Acquisition to Machine Learning

Blood vessels, spreading from the human heart to the brain and throughout the human body, constitute a significant part of the circulatory system and neural regulation. Brain tissues as well as the corresponding neural regulation and disease development rely on the normal functioning of the macrocirculation in cerebral arteries and the microcirculation in small vessel networks. In particular, the hemodynamic response to neural activation has facilitated the understanding of brain function and pathologies. Extracting cerebral vessel networks for the measurement of vascular dynamics has proved to be very important for the early diagnosis of critical cerebrovascular and neurological disorders, such as Alzheimer’s disease and stroke. To this end, extracting cerebral vessel networks for both structural and functional imaging has provided unique and complementary physiological and pathological information revealing the principles of neurovascular coupling in both healthy and diseased brain tissues. During the past 20 years, new image acquisition technologies including OCT, Photoacoustic imaging/tomography, laser speckle/Doppler imaging, functional NIRS, et al. have been developed and validated in basic neuroscience studies. At the same time, clinical contrast-enhanced imaging modalities including X-ray, CT, MRI and ultrasound have also been advanced in cerebral vessel extraction and reconstruction to provide more precise and useful diagnostic information. More than ever, machine learning for big imaging data is enabling engineers to precisely extract the cerebral vessel networks for unveiling the previously unknown information about the neurovascular coupling at a range of spatiotemporal scales.

In this Research Topic, we welcome submissions of original papers (including original research, reviews, mini-reviews) on (but not limited to) the following topics to demonstrate state-of-the-art progress in various brain vessel image acquisition (including image reconstruction) and machine learning technologies and make the convergence of both updated traditional and new imaging modalities throughout the basic research, preclinical studies and clinical applications.

1. Cutting-edge image acquisition techniques for exploring the structural and functional information about the cerebral vessel networks.
2. Various contrast-enhanced angiography techniques and machine learning in high spatiotemporal resolution for extracting the heterogeneous panorama (e.g., geometric structure, topological representation, heterogeneous signal intensity, blood flow distribution) of contrast-filled cerebral vessel networks from complex and noisy backgrounds.
3. Multitemporal and multimodal imaging for comprehensive characterization of the cerebral vasculature in the context of macro-and micro-circulation at a range of spatiotemporal scales.
4. Microscopic imaging and mathematical models of microvascular network structure for measuring the cerebral circulation and its biophysical parameters such as relative blood volume, velocity, shape and density, as well as the underlying physiological parameters such as cerebral blood flow, oxygen consumption, and rate of metabolism.
5. Various multiscale high-sensitivity brain perfusion imaging and machine learning techniques for qualitative and quantitative assessment of biophysical, physiological and pathological parameters.
6. Functional brain imaging and visualization of cerebral vessel networks for the quantitative analysis of neurovascular coupling that is spatiotemporally connecting neuronal activity to hemodynamic responses in the brain.