This repository contains an object-oriented spike sorting implementation compatible with Scikit-learn architectures as sklearn.pipeline.Pipeline and sklearn.model_selection.GridSearchCV. The classes are defined in spikesorting.py and the notebook Spikesorting_test.ipynb shows how they can be used. A Utah array recording for the test can be found in the dataset published by Brochier, et al. 2018 here. Additionally, the BlackRock library brpylib.py to read Utah array recordings can be downloaded here. The file Report.pdf gathers a detailed description of the toolbox, the context in which it was developed, and practical results of its application in data from neurorehabilitation tasks of a stroke patient.
Please write me if you have any question understanding the code ([email protected]).
- Brochier, T. et al. Data Descriptor: Massively parallel recordings in macaque motor cortex during an instructed delayed reach-to-grasp task. Sci. Data 5, (2018). doi: 10.1038/sdata.2018.55.