< Back to previous page

Publication

Toward Automated Electrode Selection in the Electronic Depth Control Strategy for Multi-unit Recordings

Book Contribution - Book Chapter Conference Contribution

Multi-electrode arrays contain an increasing number of electrodes.The manual selection of good quality signals among hundreds of electrodesbecomes impracticable for experimental neuroscientists. This increases the needfor an automated selection of electrodes containing good quality signals. Tomotivate the automated selection, three experimenters were asked to assignquality scores, taking one of four possible values, to recordings containingaction potentials obtained from the monkey primary somatosensory cortex andthe superior parietal lobule. Krippendorff’s alpha-reliability was then used toverify whether the scores, given by different experimenters, were in agreement.A Gaussian process classifier was used to automate the prediction of the signalquality using the scores of the different experimenters. Prediction accuracies ofthe Gaussian process classifier are about 80% when the quality scores ofdifferent experimenters are combined, through a median vote, to train theGaussian process classifier. It was found that predictions based also on firingrate features are in closer agreement with the experimenters’ assignments thanthose based on the signal-to-noise ratio alone.
Book: NEURAL INFORMATION PROCESSING: MODELS AND APPLICATIONS, PT II
Pages: 17 - 25
ISBN:3642175333
Publication year:2010
BOF-keylabel:yes
IOF-keylabel:yes
Authors from:Higher Education
Accessibility:Closed