NeuropixelsUltra: Dense arrays for stable, unbiased, and cell type-specific electrical imaging

点击翻译按钮获取中文摘要

Summary/Abstract Understanding the neural mechanisms underpinning cognition and behavior requires the ability to measure the dynamics and interactions of populations of neurons spread across many brain regions. Electrophysiological techniques provide the ability to measure this activity across superficial and deep structures at the speed of thought. Recent advances in electrophysiology have massively increased data quantity, quality, and ease of acquisition, thereby meaningfully reducing barriers to understanding the global brain circuits underlying behavior. A significant remaining challenge is to optimize device characteristics in order to further broaden utility, improve data quality, and accelerate the pace of research. In particular, state of the art site density is spatially too coarse to detect some cell types and neuronal processes; it remains challenging to record neurons stably in the face of brain motion; and data preprocessing is still a major limiting factor in the pace of experiments. This proposal will address these limitations by producing and evaluating a new device with >10x the number of recording sites than state-of-the-art, corresponding to an order of magnitude higher density. This device thus functions like a high-resolution electrical camera in the brain, able to image tiny electrical fields and capable of capitalizing on techniques from optics such as image registration for recording stability. We will validate and develop the new probe's characteristics by quantifying their increased ability to detect a large range of neuron types; by testing and developing their ability to track neurons across brain motion using controlled conditions; by improving algorithms towards automation of data preprocessing; and by conducting multi-modal ground-truth experiments. These probes will go beyond solving technical limitations, additionally providing new types of data: electrical imaging of `electro-morphological' shapes will enable enhanced cell-type identification and validation of neuronal biophysical models in vivo. We will disseminate the new probes, along with user-friendly software to take advantage of their improved characteristics, to `beta-tester' labs specifically interested in studying key areas of scientific opportunity. These areas include dendritic computation, freely-moving behavior, and cerebellar function, and this direct dissemination will rapidly accelerate their impact on scientific advancement.