Accurate Localization of Linear Probe Electrode Arrays across Multiple Brains.

Recently developed probes for extracellular electrophysiological recordings have large numbers of electrodes on long linear shanks. Linear electrode arrays, such as Neuropixels probes, have hundreds of recording electrodes distributed over linear shanks that span several millimeters. Because of the length of the probes, linear probe recordings in rodents usually cover multiple brain areas. Typical studies collate recordings across several recording sessions and animals. Neurons recorded in different sessions and animals thus have to be aligned to each other and to a standardized brain coordinate system. Here we evaluate two typical workflows for localization of individual electrodes in standardized coordinates. These workflows rely on imaging brains with fluorescent probe tracks and warping 3-dimensional image stacks to standardized brain atlases. One workflow is based on tissue clearing and selective plane illumination microscopy, whereas the other workflow is based on serial block-face two-photon microscopy. In both cases electrophysiological features are then used to anchor particular electrodes along the reconstructed tracks to specific locations in the brain atlas and therefore to specific brain structures. We performed groundtruth experiments, in which motor cortex outputs are labelled with ChR2 and a fluorescence protein. Light-evoked electrical activity and fluorescence can be independently localized. Recordings from brain regions targeted by the motor cortex reveal better than 0.1 mm accuracy for electrode localization, independent of workflow used. Significance statement Recent advances in silicon electrode array recording technology dramatically increased probe length and the density of electrode sites. Specifically, Neuropixels probes span multiple regions of the mouse brain across 10 millimeter long shanks. To localize recorded neurons, it is critical to localize recording sites. Here we develop two workflows based on brain-wide imaging of probe tracks and analysis of electrophysiological landmarks in a standardized brain coordinate system. One workflow is based on lightsheet microscopy. The other workflow involves serial block-face two-photon microscopy. We evaluate these workflows based on ground truth experiments. We show that electrodes, and thereby recorded neurons, can be localized to better than 100 micrometers.