neuropixels-analysis

You can run spike sorting on Neuropixels data by loading raw SpikeGLX or Open Ephys recordings into SpikeInterface to preprocess, apply drift correction, execute Kilosort4, compute quality metrics, and curate units.

AI-assisted curation analyzes spike sorting outputs visually to help extract clean, well-isolated neural units. It evaluates quality metrics to distinguish real units from noise before exporting to Phy or NWB.

Yes, the SpikeInterface pipeline supports both NP1.0 and NP2.0 probes. It handles preprocessing, motion correction, and Kilosort4 spike sorting across these Neuropixels probe types to extract well-isolated units.

Yes, you can export curated spike sorting results to Phy and NWB formats. After AI-assisted curation and quality metric computation, the pipeline allows exporting clean units for downstream visualization and analysis.

The best way to handle drift correction in Neuropixels recordings is using SpikeInterface's motion correction module during preprocessing. This stabilizes the data spatially before applying Kilosort4 for spike sorting.

You need Kilosort4 for spike sorting, but quality metrics and curation are handled separately. SpikeInterface computes quality metrics on the sorting output, followed by AI-assisted visual analysis to extract clean units.