1038-NeuroXiv: AI-Powered Brain-Wide Neuron Morphometry

This paper introduces NeuroXiv, a sophisticated open-access database designed to standardize and host over 175,000 reconstructed neuron morphologies within a unified brain atlas. It addresses existing gaps in neuroinformatics by aggregating diverse datasets into a single, scalable platform that maps individual neurons to the Common Coordinate Framework. A core feature of this resource is the AI-Powered Open Mining (AIPOM) engine, which allows researchers to perform complex data searches using natural language queries. The platform utilizes a Mixture of Experts (MoE) framework to generate accurate, automated reports on cell types, connectivity patterns, and morphological features. Furthermore, NeuroXiv offers interactive 3D visualization tools and optimized client-side processing to ensure rapid analysis of large-scale anatomical data. Ultimately, this resource streamlines the study of brain-wide neural networks by making high-resolution morphometry more accessible and reusable for the global scientific community.References:Jiang S, Wang L, Yun Z, et al. NeuroXiv: AI-powered open databasing and dynamic mining of brain-wide neuron morphometry[J]. Nature methods, 2025, 22(6): 1195-1198.前往小宇宙评论区与主播互动