Whole Brain Vessel Graphs: A Dataset and Benchmark for Graph Learning and Neuroscience (VesselGraph) (2108.13233v2)

Abstract: Biological neural networks define the brain function and intelligence of humans and other mammals, and form ultra-large, spatial, structured graphs. Their neuronal organization is closely interconnected with the spatial organization of the brain's microvasculature, which supplies oxygen to the neurons and builds a complementary spatial graph. This vasculature (or the vessel structure) plays an important role in neuroscience; for example, the organization of (and changes to) vessel structure can represent early signs of various pathologies, e.g. Alzheimer's disease or stroke. Recently, advances in tissue clearing have enabled whole brain imaging and segmentation of the entirety of the mouse brain's vasculature. Building on these advances in imaging, we are presenting an extendable dataset of whole-brain vessel graphs based on specific imaging protocols. Specifically, we extract vascular graphs using a refined graph extraction scheme leveraging the volume rendering engine Voreen and provide them in an accessible and adaptable form through the OGB and PyTorch Geometric dataloaders. Moreover, we benchmark numerous state-of-the-art graph learning algorithms on the biologically relevant tasks of vessel prediction and vessel classification using the introduced vessel graph dataset. Our work paves a path towards advancing graph learning research into the field of neuroscience. Complementarily, the presented dataset raises challenging graph learning research questions for the machine learning community, in terms of incorporating biological priors into learning algorithms, or in scaling these algorithms to handle sparse,spatial graphs with millions of nodes and edges. All datasets and code are available for download at https://github.com/jocpae/VesselGraph .

• The paper presents VesselGraph, an extendable dataset that enables graph learning on murine brain vasculature.
• The methodology integrates advanced imaging and graph extraction techniques with algorithms like SEAL, achieving a ROC AUC >90% in link prediction.
• Experimental results validate the dataset's utility for benchmarking vessel prediction and node classification, paving the way for improved biomedical models.

An Overview of "Whole Brain Vessel Graphs: A Dataset and Benchmark for Graph Learning and Neuroscience"

The research paper "Whole Brain Vessel Graphs: A Dataset and Benchmark for Graph Learning and Neuroscience (VesselGraph)" offers a profound contribution to the intersection of neuroscience and graph learning. This paper introduces an extendable dataset for graph learning, particularly focusing on the vascular structure of the murine brain. The significance of this work lies in the integration of advanced graph learning methodologies with the intricate challenges posed by the brain's vascular connectome.

Dataset and Methodology

The paper constructs a dataset of whole-brain vessel graphs using sophisticated imaging protocols and graph extraction techniques. The data acquisition leverages methods such as the Voreen volume rendering engine and various microscopy techniques, culminating in an accessible dataset amenable to graph learning frameworks like OGB and PyTorch Geometric. This represents a step forward in processing spatially structured biological data, where the vascular graphs are provided in forms conducive to computational manipulation.

Graph Learning Benchmarks

The paper benchmarks state-of-the-art graph learning algorithms on two notable tasks: vessel prediction (link prediction) and vessel classification (node classification). These tasks are pivotal as they address the structural integrity and biological classification of brain vasculature. The authors implement various learning techniques, including GCN, GraphSAGE, and SEAL, demonstrating a spectrum of performance levels. Notably, the SEAL algorithm, with its labeling trick, achieved superior results in link prediction tasks, highlighting the utility of sophisticated graph neural network architectures for spatially dependent data.

Experimental Results

The experiment results underscore the diversified performance across algorithms, with the SEAL model yielding a ROC AUC > 90% in link prediction, demonstrating its robustness in handling the complexity of spatial graphs. In node classification, techniques like GraphSAGE and Cluster-GCN performed well, with balanced accuracy metrics revealing their capacity to accommodate class imbalances inherent in biological datasets. These experiments not only validate model effectiveness but also emphasize the necessity for tailored methodologies to capture the biological relevance of such data.

Implications and Future Directions

This dataset and its analysis pave the way for more profound integration of graph learning in biological and medical research. The availability of a large-scale, intricately-detailed dataset facilitates the exploration of graph learning algorithms’ capabilities in understanding neural and vascular interactions. The VesselGraph dataset encourages exploration into incorporating biological priors into learning frameworks and scaling these to efficiently handle vast, sparse networks.

Theoretically, this work stimulates discussions on the possibility of improved biomedical models and simulations, particularly for pathologies involving vascular anomalies. Practically, it presents opportunities for developing diagnostic tools and advancing our comprehension of cerebrovascular biology.

As the paper acknowledges, the limitations imposed by technical imaging constraints and the inherent biases within the dataset highlight the need for continuous enhancements in imaging technologies and graph representation techniques. Future research could expand on the current dataset by incorporating heterogeneous graph representations or incorporating additional features to provide a holistic biological insight.

In summary, this paper lays foundational work for the merger of computational graph learning techniques with neuroscience, opening avenues for future research to unravel the complexities of brain function and disorders through the structure of its vessels. The VesselGraph dataset stands as a benchmark facilitating this interdisciplinary exploration, holding immense potential for scientific advancements in understanding the brain's functional mechanics intertwined with its vascular architecture.