Limits...
Automatic online spike sorting with singular value decomposition and fuzzy C-mean clustering.

Oliynyk A, Bonifazzi C, Montani F, Fadiga L - BMC Neurosci (2012)

Bottom Line: The results of these tests showed an excellent accuracy in discriminating low-amplitude and overlapping spikes under strong background noise.This new software provides neuroscience laboratories with a new tool for fast and robust online classification of single neuron activity.This feature could become crucial in situations when online spike detection from multiple electrodes is paramount, such as in human clinical recordings or in brain-computer interfaces.

View Article: PubMed Central - HTML - PubMed

Affiliation: Section of Human Physiology, Department of Biomedical Sciences and Advanced Therapies, Faculty of Medicine, University of Ferrara, Via Fossato di Mortara 17/19, 44121, Ferrara, Italy. lynnry@unife.it

ABSTRACT

Background: Understanding how neurons contribute to perception, motor functions and cognition requires the reliable detection of spiking activity of individual neurons during a number of different experimental conditions. An important problem in computational neuroscience is thus to develop algorithms to automatically detect and sort the spiking activity of individual neurons from extracellular recordings. While many algorithms for spike sorting exist, the problem of accurate and fast online sorting still remains a challenging issue.

Results: Here we present a novel software tool, called FSPS (Fuzzy SPike Sorting), which is designed to optimize: (i) fast and accurate detection, (ii) offline sorting and (iii) online classification of neuronal spikes with very limited or human intervention. The method is based on a combination of Singular Value Decomposition for fast and highly accurate pre-processing of spike shapes, unsupervised Fuzzy C-mean, high-resolution alignment of extracted spike waveforms, optimal selection of the number of features to retain, automatic identification the number of clusters, and quantitative quality assessment of resulting clusters independent on their size. After being trained on a short testing data stream, the method can reliably perform supervised online classification and monitoring of single neuron activity. The generalized procedure has been implemented in our FSPS spike sorting software (available free for non-commercial academic applications at the address: http://www.spikesorting.com) using LabVIEW (National Instruments, USA). We evaluated the performance of our algorithm both on benchmark simulated datasets with different levels of background noise and on real extracellular recordings from premotor cortex of Macaque monkeys. The results of these tests showed an excellent accuracy in discriminating low-amplitude and overlapping spikes under strong background noise. The performance of our method is competitive with respect to other robust spike sorting algorithms.

Conclusions: This new software provides neuroscience laboratories with a new tool for fast and robust online classification of single neuron activity. This feature could become crucial in situations when online spike detection from multiple electrodes is paramount, such as in human clinical recordings or in brain-computer interfaces.

Show MeSH

Related in: MedlinePlus

Fully automated online classifier.A - After the period of “TEST” acquisition, which starts once for each recording site, the classification of newly arriving spikes is continual. The control panel on the left-hand side permits some adjustments to automatic thresholding and spline interpolation parameters, depending on the digitalization rate of acquisition. Numerical information about the number of spikes and single units, as well as their waveforms, are instantly available to the researcher; B – interface window of online monitoring. This is the most innovative part of our procedure. To ensure good performance we gave the option of sharing most resource-dependent processes, like extraction of PCs and their FCM-classification, between two different processors. In order to achieve this goal we created two separated subprograms (VI 1 and VI 2), running in parallel and linked via UDP protocol, for the transmission of the reduced number of extracted features (PCs) together with a time of spike occurrence, which is reverse-reconstructed and reproduced right after event classification. These two VIs, being processor-dependent, can be run on the same computer when a limited number of recording electrodes is considered. Since the latest versions of LabVIEW (LabVIEW-2009 or higher) can effectively treat multi-core processor architecture and parallel-loop execution, the FSPS software can run on the same computer, sharing the power of multi-core processor (Intel Core i5-2430 M, 2.4 GHz). However, an Ethernet connection may also be useful when experimental conditions necessitate distant online monitoring.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3473300&req=5

Figure 3: Fully automated online classifier.A - After the period of “TEST” acquisition, which starts once for each recording site, the classification of newly arriving spikes is continual. The control panel on the left-hand side permits some adjustments to automatic thresholding and spline interpolation parameters, depending on the digitalization rate of acquisition. Numerical information about the number of spikes and single units, as well as their waveforms, are instantly available to the researcher; B – interface window of online monitoring. This is the most innovative part of our procedure. To ensure good performance we gave the option of sharing most resource-dependent processes, like extraction of PCs and their FCM-classification, between two different processors. In order to achieve this goal we created two separated subprograms (VI 1 and VI 2), running in parallel and linked via UDP protocol, for the transmission of the reduced number of extracted features (PCs) together with a time of spike occurrence, which is reverse-reconstructed and reproduced right after event classification. These two VIs, being processor-dependent, can be run on the same computer when a limited number of recording electrodes is considered. Since the latest versions of LabVIEW (LabVIEW-2009 or higher) can effectively treat multi-core processor architecture and parallel-loop execution, the FSPS software can run on the same computer, sharing the power of multi-core processor (Intel Core i5-2430 M, 2.4 GHz). However, an Ethernet connection may also be useful when experimental conditions necessitate distant online monitoring.

Mentions: The basic strategy of FSPS software is to provide accurate and trustable classification with a minimal supervision and, more importantly, without specific software knowledge (like Python scripting, Matlab toolboxes, C++ etc.). The software supports a large variety of digital acquisition (DAQ) systems (including low-cost ones) and simplifies electrophysiology setup by using the flexible graphical user interface (GUI) of Virtual Instruments (VIs). The spike sorting algorithm was entirely implemented within graphical programming language LabVIEW 2009 (National Instruments, USA), whose DAQ hardware and interfaces became very popular in electrophysiology labs over the last decade. Besides, we choose this software platform for its ability to control the experimental protocol and data acquisition while being able to run the analysis fast and online using threaded dataflow methodology[21]. It is also reported that many LabVIEW subroutines shows considerable outperformance when compared to their identical counterpart written in MATLAB (MathWorks, USA)[22]. FSPS high-level schema is sketched in Figure 1. The program allows triggered (Figure 2) and continuous (Figure 3A) acquisition from one or more electrodes simultaneously and offers the user the choice to set all parameters of the analysis automatically or manually, both in case of “test” acquisition and online classification (Figure 3B). Besides, the software has the following advanced features: band-pass signal filtering; automatic detection of spikes with evaluation of background noise level and automatic threshold selection; extraction and alignment of spike waveforms; removal of constant DC offset, false positive and noisy spikes; pre-processing with computationally efficient PCA; automatic determination the number of PCs to retain; automatic determination of the number of clusters to be found; offline fuzzy clustering analysis; online fuzzy classification; 2D and 3D visualization tools; quantitative quality assessment of resulting clusters, basic statistics, PSTH, measurements of some clinical parameters of spike trains etc. Additional file1. The software allows simultaneous visualization/monitoring of activity of several isolated neurons and provides online acoustic feedback about one selected neuron. It has import/export features and allows synchronization of the acquisition with external devices (e.g. digital videorecorders, stimulators etc.). The application is available athttp://www.spikesorting.com in the Download section.


Automatic online spike sorting with singular value decomposition and fuzzy C-mean clustering.

Oliynyk A, Bonifazzi C, Montani F, Fadiga L - BMC Neurosci (2012)

Fully automated online classifier.A - After the period of “TEST” acquisition, which starts once for each recording site, the classification of newly arriving spikes is continual. The control panel on the left-hand side permits some adjustments to automatic thresholding and spline interpolation parameters, depending on the digitalization rate of acquisition. Numerical information about the number of spikes and single units, as well as their waveforms, are instantly available to the researcher; B – interface window of online monitoring. This is the most innovative part of our procedure. To ensure good performance we gave the option of sharing most resource-dependent processes, like extraction of PCs and their FCM-classification, between two different processors. In order to achieve this goal we created two separated subprograms (VI 1 and VI 2), running in parallel and linked via UDP protocol, for the transmission of the reduced number of extracted features (PCs) together with a time of spike occurrence, which is reverse-reconstructed and reproduced right after event classification. These two VIs, being processor-dependent, can be run on the same computer when a limited number of recording electrodes is considered. Since the latest versions of LabVIEW (LabVIEW-2009 or higher) can effectively treat multi-core processor architecture and parallel-loop execution, the FSPS software can run on the same computer, sharing the power of multi-core processor (Intel Core i5-2430 M, 2.4 GHz). However, an Ethernet connection may also be useful when experimental conditions necessitate distant online monitoring.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3473300&req=5

Figure 3: Fully automated online classifier.A - After the period of “TEST” acquisition, which starts once for each recording site, the classification of newly arriving spikes is continual. The control panel on the left-hand side permits some adjustments to automatic thresholding and spline interpolation parameters, depending on the digitalization rate of acquisition. Numerical information about the number of spikes and single units, as well as their waveforms, are instantly available to the researcher; B – interface window of online monitoring. This is the most innovative part of our procedure. To ensure good performance we gave the option of sharing most resource-dependent processes, like extraction of PCs and their FCM-classification, between two different processors. In order to achieve this goal we created two separated subprograms (VI 1 and VI 2), running in parallel and linked via UDP protocol, for the transmission of the reduced number of extracted features (PCs) together with a time of spike occurrence, which is reverse-reconstructed and reproduced right after event classification. These two VIs, being processor-dependent, can be run on the same computer when a limited number of recording electrodes is considered. Since the latest versions of LabVIEW (LabVIEW-2009 or higher) can effectively treat multi-core processor architecture and parallel-loop execution, the FSPS software can run on the same computer, sharing the power of multi-core processor (Intel Core i5-2430 M, 2.4 GHz). However, an Ethernet connection may also be useful when experimental conditions necessitate distant online monitoring.
Mentions: The basic strategy of FSPS software is to provide accurate and trustable classification with a minimal supervision and, more importantly, without specific software knowledge (like Python scripting, Matlab toolboxes, C++ etc.). The software supports a large variety of digital acquisition (DAQ) systems (including low-cost ones) and simplifies electrophysiology setup by using the flexible graphical user interface (GUI) of Virtual Instruments (VIs). The spike sorting algorithm was entirely implemented within graphical programming language LabVIEW 2009 (National Instruments, USA), whose DAQ hardware and interfaces became very popular in electrophysiology labs over the last decade. Besides, we choose this software platform for its ability to control the experimental protocol and data acquisition while being able to run the analysis fast and online using threaded dataflow methodology[21]. It is also reported that many LabVIEW subroutines shows considerable outperformance when compared to their identical counterpart written in MATLAB (MathWorks, USA)[22]. FSPS high-level schema is sketched in Figure 1. The program allows triggered (Figure 2) and continuous (Figure 3A) acquisition from one or more electrodes simultaneously and offers the user the choice to set all parameters of the analysis automatically or manually, both in case of “test” acquisition and online classification (Figure 3B). Besides, the software has the following advanced features: band-pass signal filtering; automatic detection of spikes with evaluation of background noise level and automatic threshold selection; extraction and alignment of spike waveforms; removal of constant DC offset, false positive and noisy spikes; pre-processing with computationally efficient PCA; automatic determination the number of PCs to retain; automatic determination of the number of clusters to be found; offline fuzzy clustering analysis; online fuzzy classification; 2D and 3D visualization tools; quantitative quality assessment of resulting clusters, basic statistics, PSTH, measurements of some clinical parameters of spike trains etc. Additional file1. The software allows simultaneous visualization/monitoring of activity of several isolated neurons and provides online acoustic feedback about one selected neuron. It has import/export features and allows synchronization of the acquisition with external devices (e.g. digital videorecorders, stimulators etc.). The application is available athttp://www.spikesorting.com in the Download section.

Bottom Line: The results of these tests showed an excellent accuracy in discriminating low-amplitude and overlapping spikes under strong background noise.This new software provides neuroscience laboratories with a new tool for fast and robust online classification of single neuron activity.This feature could become crucial in situations when online spike detection from multiple electrodes is paramount, such as in human clinical recordings or in brain-computer interfaces.

View Article: PubMed Central - HTML - PubMed

Affiliation: Section of Human Physiology, Department of Biomedical Sciences and Advanced Therapies, Faculty of Medicine, University of Ferrara, Via Fossato di Mortara 17/19, 44121, Ferrara, Italy. lynnry@unife.it

ABSTRACT

Background: Understanding how neurons contribute to perception, motor functions and cognition requires the reliable detection of spiking activity of individual neurons during a number of different experimental conditions. An important problem in computational neuroscience is thus to develop algorithms to automatically detect and sort the spiking activity of individual neurons from extracellular recordings. While many algorithms for spike sorting exist, the problem of accurate and fast online sorting still remains a challenging issue.

Results: Here we present a novel software tool, called FSPS (Fuzzy SPike Sorting), which is designed to optimize: (i) fast and accurate detection, (ii) offline sorting and (iii) online classification of neuronal spikes with very limited or human intervention. The method is based on a combination of Singular Value Decomposition for fast and highly accurate pre-processing of spike shapes, unsupervised Fuzzy C-mean, high-resolution alignment of extracted spike waveforms, optimal selection of the number of features to retain, automatic identification the number of clusters, and quantitative quality assessment of resulting clusters independent on their size. After being trained on a short testing data stream, the method can reliably perform supervised online classification and monitoring of single neuron activity. The generalized procedure has been implemented in our FSPS spike sorting software (available free for non-commercial academic applications at the address: http://www.spikesorting.com) using LabVIEW (National Instruments, USA). We evaluated the performance of our algorithm both on benchmark simulated datasets with different levels of background noise and on real extracellular recordings from premotor cortex of Macaque monkeys. The results of these tests showed an excellent accuracy in discriminating low-amplitude and overlapping spikes under strong background noise. The performance of our method is competitive with respect to other robust spike sorting algorithms.

Conclusions: This new software provides neuroscience laboratories with a new tool for fast and robust online classification of single neuron activity. This feature could become crucial in situations when online spike detection from multiple electrodes is paramount, such as in human clinical recordings or in brain-computer interfaces.

Show MeSH
Related in: MedlinePlus