Unsupervised Detection of Cell Assemblies with Graph Neural Networks

Roman Koshkin, Tomoki Fukai

Keywords: graph neural networks, cell assemblies, neural patterns

TL;DR: Graph neural networks hold promise for unsupervised detection of patterned neural activity.

Abstract: Cell assemblies, putative units of neural computation, manifest themselves as repeating and temporally coordinated activity of neurons. However, understanding of their role in brain function is hampered by a lack of scalable methods for their unsupervised detection. We propose using a graph neural network for embedding spike data into a sequence of fixed size vectors and clustering them based on their self-similarity across time. We validate our method on synthetic data and real neural recordings.