Unsupervised wavelet-based spike sorting with dynamic codebook searching and replenishment

  • Authors:

  • Hsiao-Lung Chan;Tony Wu;Shih-Tseng Lee;Ming-An Lin;Shau-Ming He;Pei-Kuang Chao;Yu-Tai Tsai

  • Affiliations:

  • Department of Electrical Engineering, Chang Gung University, 259 Wenhwa 1st Road, Kweishan, Taoyuan 333, Taiwan and Medical Augmented Reality Research Center, Chang Gung University, Taoyuan, Taiwa ...;Department of Neurology, Chang Gung Memorial Hospital, Taoyuan, Taiwan and Medical Augmented Reality Research Center, Chang Gung University, Taoyuan, Taiwan;Department of Neurosurgery, Chang Gung Memorial Hospital, Taoyuan, Taiwan and Medical Augmented Reality Research Center, Chang Gung University, Taoyuan, Taiwan;Department of Electrical Engineering, Chang Gung University, 259 Wenhwa 1st Road, Kweishan, Taoyuan 333, Taiwan;Department of Electrical Engineering, Chang Gung University, 259 Wenhwa 1st Road, Kweishan, Taoyuan 333, Taiwan;Department of Electrical Engineering, Chang Gung University, 259 Wenhwa 1st Road, Kweishan, Taoyuan 333, Taiwan;Department of Electrical Engineering, Chang Gung University, 259 Wenhwa 1st Road, Kweishan, Taoyuan 333, Taiwan and Department of Neurology, Chang Gung Memorial Hospital, Taoyuan, Taiwan

  • Venue:
  • Neurocomputing
  • Year:
  • 2010

Quantified Score

Hi-index 0.01

Visualization

Abstract

Spike information is beneficial for correlating neuronal activity to various stimuli, finding target neural areas for deep brain stimulation, and decoding intended motor command for brain-machine interface. Unsupervised classification based on spike features provides a way to separate spikes generated from different neurons. Here, we propose an unsupervised spike sorting method based on specific wavelet coefficients (SWC) and using both a new spike alignment technique based on multi-peak energy comparison (MPEC) and a dynamic codebook-based template-matching algorithm with a class-merging feature. The MPEC alignment reduced inconsistent alignment caused by spike deformation. Using SWC not only reduced the number of features but also performed better in terms of matching a neuronal spike to its own class than relying on spike waveform or whole wavelet coefficients. Moreover, the employed codebook searching and replenishment can be operated in an online, real-time mode.