This title appears in the Scientific Report :
2019
Please use the identifier:
http://hdl.handle.net/2128/23510 in citations.
Detection and Removal of Artefacts in Multi-Channel Electrophysiology Recordings
Detection and Removal of Artefacts in Multi-Channel Electrophysiology Recordings
Modern electrophysiological experiments using multi-electrode arrays enable simultaneous access to the spiking activity of more than one hundred single neurons. However, these recordings reveal types of artefacts that have likely been overseen in conventional recordings with a small number of electr...
Saved in:
Personal Name(s): | Essink, Simon (Corresponding author) |
---|---|
Kleinjohann, Alexander (Corresponding author) / Barthélemy, Frédéric / Ito, Junji / Riehle, Alexa / Grün, Sonja / Brochier, Thomas | |
Contributing Institute: |
Jara-Institut Brain structure-function relationships; INM-10 Computational and Systems Neuroscience; IAS-6 Computational and Systems Neuroscience; INM-6 |
Imprint: |
2019
|
Conference: | Bernstein Conference 2019, Berlin (Germany), 2019-09-17 - 2019-09-20 |
Document Type: |
Poster |
Research Program: |
GRK 2416: MultiSenses-MultiScales: Neue Ansätze zur Aufklärung neuronaler multisensorischer Integration Human Brain Project Specific Grant Agreement 2 Doktorand ohne besondere Förderung Connectivity and Activity |
Link: |
OpenAccess OpenAccess |
Publikationsportal JuSER |
Modern electrophysiological experiments using multi-electrode arrays enable simultaneous access to the spiking activity of more than one hundred single neurons. However, these recordings reveal types of artefacts that have likely been overseen in conventional recordings with a small number of electrodes. One signature of such artefacts are hyper-synchronous putative spikes at sampling rate precision (synchrofacts [1]), sometimes involving a large number of channels, which are unlikely to represent neuronal activity. Detection and removal of synchrofacts can be performed on the spike-sorted data [2], but here we examine the raw signals for signatures of synchrofacts, aiming at artefact removal on these signals before spike-sorting.We currently explore datasets from a visuo-motor tracking experiment in monkeys [3] with four Utah arrays of 36 electrodes each inserted in V1, V2, DP and area 7A, and one array of 100 electrodes in M1/PMd. These recordings provide 224 active channels of raw signals (sampled at 30 kHz, filtered between 0.3 and 7500 Hz) from which we extract putative spikes by thresholding the high pass filtered (>250 Hz) raw signal. Signals passing the threshold are then spike-sorted into the activity of (multiple) single neurons recorded on each channel (~200 single units in total).We searched for causes of the synchrofacts in the raw data in two ways: 1.) based on pairwise correlation analysis of any two channels, and 2.) based on the channel-averaged signal. The correlation-based analysis revealed that about 10% of channels show high pairwise correlations across frequency bands which directly cause synchrofacts. Furthermore, groups of channels are correlated in narrow frequency bands. The analysis of the channel-averaged signal showed low amplitude negative deflections at synchrofact times. The distribution of widths of these deflections is bi-modal indicating two distinct sources.We are able to suppress 2/3 of the synchrofacts in the data by removing cross-talking channels and the negative deflections from the raw signals. Remaining synchrofacts might be related to narrow-band oscillations, which we aim to remove by using the Joint Decorrelation method [4].As the detection of artefacts can be performed on the raw signals, we propose to clean the data before spike-sorting. We plan to integrate this procedure in a standardized and reproducible pre-processing workflow and compare the results of spike-sorting based on raw versus cleaned data.AcknowledgementsWe thank Bejamin Dann for valuable discussions. This project has received funding from Associated International Laboratory (LIA) between Research Center Jülich and INT, RTG2416 MultiSenses-MultiScales (Deutsche Forschungsgemeinschaft) and EU Grant 785907 (HBP).References Sprenger, J. Spatial Dependence of the Spike-Related Component of the Local Field Potential in Motor Cortex (Master’s thesis, RWTH Aachen) (2014). Torre et al. Synchronous Spike Patterns in Macaque Motor Cortex during an Instructed-Delay Reach-to-Grasp Task. J. Neurosci. 36(32):8329-8340 (2016)., 10.1523/JNEUROSCI.4375-15.2016 de Haan, M. J., Brochier, T., Grün, S., Riehle, A. & Barthélemy, F. V. Real-time visuomotor behavior and electrophysiology recording setup for use with humans and monkeys. J. Neurophysiol. 120, 539–552 (2018)., 10.1152/jn.00262.2017 de Cheveigné A, Parra LC. Joint decorrelation, a versatile tool for multichannel data analysis. Neuroimage 98:487-505 (2014)., 10.1016/j.neuroimage.2014.05.068 |