Mapping brain-wide activity networks: brainways as a tool for neurobiological discovery.

Abstract

Identifying brain-wide neural circuits and targeting these areas for neuropharmacological interventions are significant challenges in contemporary neuroscience. Traditional methods for registering and quantifying fluorescence in brain slices are labor-intensive and struggle to extract functional insights from complex datasets. To address these challenges, we introduce Brainways-an AI-based, open-source software that streamlines neural network identification from digital imaging to network analysis. Brainways facilitates neurobiological research by enabling automatic registration of coronal brain slices to any 3D brain atlas, along with precise quantification of fluorescent markers, such as activity markers and tracers, across brain regions. Brainways incorporates advanced statistical tools to identify neural patterns and functional networks associated with specific experimental contrasts. Trained on rat and mouse brain atlases, Brainways achieves over 93% atlas registration accuracy. The software also allows users to easily adjust the automatic registration through a user-friendly interface for enhanced accuracy. We present two experiment analyses demonstrating Brainways' capabilities. The first replicates and extends findings from a prior experiment on pro-social behavior in rats, wherein rats learned to free a trapped cagemate from a restrainer under ingroup and outgroup social conditions. Using Brainways, we analyzed approximately 300 times more tissue area than in our previous manual approach. The second experiment utilizes Multiplex RNAscope imaging for whole-brain registration, enabling combined quantification of cell type expression and activity markers. These analyses highlight Brainways' ability to link specific cell types and their activity to task conditions, providing detailed neural insights. Brainways offers a rapid and accurate solution for large-scale neurobiological projects, creating new opportunities to understand neural networks underlying complex behaviors.