Hierarchical Adaptive Means (HAM) clustering for hardware-efficient, unsupervised and real-time spike sorting

The neural command from motor neurons to muscles -sometimes referred to as the neural drive to muscle -can be identified by decomposition of electromyographic (EMG) paper proposes for the first time an innovative method for fully automatic and real-time intramuscular EMG (iEMG) decomposition. The method is based on online single-pass density-based clustering and adaptive classification of bivariate features, using the concept of potential measure. No attempt was made to resolve superimposed motor unit action potentials. The proposed algorithm was validated on sets of simulated and experimental iEMG signals. Signals were recorded from the biceps femoris long-head, vastus medialis and lateralis and tibialis anterior muscles during low-to-moderate isometric constant-force and linearly-varying force contractions. The average number of missed, duplicated and erroneous clusters for the examined signals was 0.5 ± 0.8, 1.2 ± 1.0, and 1.0 ± 0.8, respectively. The average decomposition accuracy (defined similar to signal detection theory but without using True Negatives in the denominator) and coefficient of determination (variance accounted for) for the cumulative discharge rate estimation were 70 ± 9%, and 94 ± 5%, respectively. The time cost for processing each 200 ms iEMG interval was 43 ± 16 (21-97) ms. However, computational time generally increases over time as a function of frames/signal epochs. Meanwhile, the incremental accuracy defined as the accuracy of real-time analysis of each signal epoch, was 74 ± 18% for epochs recorded after initial one second. The proposed algorithm is thus a promising new tool for neural decoding in the next-generation of prosthetic control.

show abstract

“…59 The definition of IA is the same as in Eq. (11) but it is calculated in each frame during the algorithm progression.…”

mentioning

confidence: 99%

A Real-Time Method for Decoding the Neural Drive to Muscles Using Single-Channel Intra-Muscular EMG Recordings

Karimimehr

Marateb

Muceli

et al. 2017

Int. J. Neur. Syst.

View full text Add to dashboard Cite

show abstract

“…These algorithms are constantly evolving, highlighting the importance of the problem, the need for accurate detection automats, and the complexity in identifying neurons recorded from the extracellular space (Azami et al, 2015;Franke et al, 2015;Kaneko et al, 2007;Paraskevopoulou et al, 2014;Rall, 1962).…”

Section: Sensors Recording Neuronal Activitymentioning

confidence: 99%

Intracranial neuronal ensemble recordings and analysis in epilepsy

Tóth

Fabó

Entz

et al. 2016

Journal of Neuroscience Methods

View full text Add to dashboard Cite

Pathological neuronal firing was demonstrated 50 years ago as the hallmark of epileptically transformed cortex with the use of implanted microelectrodes. Since then, microelectrodes remained only experimental tools in humans to detect unitary neuronal activity to reveal physiological and pathological brain functions. This recording technique has evolved substantially in the past few decades; however, based on recent human data implying their usefulness as diagnostic tools, we expect a substantial increase in the development of microelectrodes in the near future.Here, we review the technological background and history of microelectrode array development for human examinations in epilepsy, including discussions on of wire-based and microelectrode arrays fabricated using micro-electro-mechanical system (MEMS) techniques and novel future techniques to record neuronal ensemble.We give an overview of clinical and surgical considerations, and try to provide a list of probes on the market with their availability for human recording. 2Then finally, we briefly review the literature on modulation of single neuron for the treatment of epilepsy, and highlight the current topics under examination that can be background for the future development.3

show abstract

“…The clustering can be performed in a variety of feature spaces spanned by features such as peak or valley amplitude, principal components, or wavelet coefficients. In addition to these commercial or widely used tools, new algorithms for performing spike sorting continue to be developed [5][6][7][8][9][10][11][12].…”

Section: Introductionmentioning

confidence: 99%

Taller Peaks: an improved spike detection algorithm that simultaneously reduces type I and type II errors for Wave{\_}clus

Okatan¹

2017

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

Abstract:Wave clus is an unsupervised spike detection and sorting algorithm that has been used in dozens of experimental studies as a spike sorting tool. It is often used as a benchmark for comparing the performance of new spike sorting algorithms. For these reasons, the spike detection performance of Wave clus is important for both experimental and computational studies that involve spike sorting. Two measures of spike detection performance are the number of false positive detections (type I error) and the number of missed spikes (type II error). Here, a new spike detection algorithm is proposed that reduces the number of misses and false positives of Wave clus in a widely used simulated data set across the entire range of commonly used detection thresholds. The algorithm accepts a spike if its amplitude is larger than the amplitude of its two immediate neighbors, where an immediate neighbor is the nearest peak of the same polarity within ± 1 refractory period. The simultaneous reduction that is achieved in the number of false positives and misses is important for experimental and computational studies that use Wave clus as a spike sorting tool or as a benchmark. A software patch that incorporates the algorithm into Wave clus as an optional spike detection algorithm is provided.

show abstract

Hierarchical Adaptive Means (HAM) clustering for hardware-efficient, unsupervised and real-time spike sorting

Cited by 28 publications

References 37 publications

A Real-Time Method for Decoding the Neural Drive to Muscles Using Single-Channel Intra-Muscular EMG Recordings

A Real-Time Method for Decoding the Neural Drive to Muscles Using Single-Channel Intra-Muscular EMG Recordings

Intracranial neuronal ensemble recordings and analysis in epilepsy

Taller Peaks: an improved spike detection algorithm that simultaneously reduces type I and type II errors for Wave{\_}clus

Contact Info

Product

Resources

About