Limits...
Spike sorting of heterogeneous neuron types by multimodality-weighted PCA and explicit robust variational Bayes.

Takekawa T, Isomura Y, Fukai T - Front Neuroinform (2012)

Bottom Line: The performance of the proposed method was compared with that of other conventional methods for simulated and experimental data sets.We found that the mPCA efficiently extracts highly informative features as clusters clearly separable in a relatively low-dimensional feature space.Our method showed significantly improved performance in spike sorting of these "difficult" neurons.

View Article: PubMed Central - PubMed

Affiliation: Laboratory for Neural Circuit Theory, RIKEN Brain Science Institute Wako, Japan.

ABSTRACT
This study introduces a new spike sorting method that classifies spike waveforms from multiunit recordings into spike trains of individual neurons. In particular, we develop a method to sort a spike mixture generated by a heterogeneous neural population. Such a spike sorting has a significant practical value, but was previously difficult. The method combines a feature extraction method, which we may term "multimodality-weighted principal component analysis" (mPCA), and a clustering method by variational Bayes for Student's t mixture model (SVB). The performance of the proposed method was compared with that of other conventional methods for simulated and experimental data sets. We found that the mPCA efficiently extracts highly informative features as clusters clearly separable in a relatively low-dimensional feature space. The SVB was implemented explicitly without relying on Maximum-A-Posterior (MAP) inference for the "degree of freedom" parameters. The explicit SVB is faster than the conventional SVB derived with MAP inference and works more reliably over various data sets that include spiking patterns difficult to sort. For instance, spikes of a single bursting neuron may be separated incorrectly into multiple clusters, whereas those of a sparsely firing neuron tend to be merged into clusters for other neurons. Our method showed significantly improved performance in spike sorting of these "difficult" neurons. A parallelized implementation of the proposed algorithm (EToS version 3) is available as open-source code at http://etos.sourceforge.net/.

No MeSH data available.


Typical examples of the posterior distributions of the DOF parameters for the clusters estimated by MAP-SVB and explicit SVB. For MAP-SVB, the vertical lines indicate the values that maximize the distributions. These values are used for the model estimation. For explicit SVB, the vertical lines indicate the average values of the distributions, and explicit SVB takes into account the shapes of the posterior distributions for the model estimation.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3307250&req=5

Figure 7: Typical examples of the posterior distributions of the DOF parameters for the clusters estimated by MAP-SVB and explicit SVB. For MAP-SVB, the vertical lines indicate the values that maximize the distributions. These values are used for the model estimation. For explicit SVB, the vertical lines indicate the average values of the distributions, and explicit SVB takes into account the shapes of the posterior distributions for the model estimation.

Mentions: To explain why explicit SVB is advantageous over MAP-SVB, in Figure 7 we display the posterior distributions of the DOF parameters for the clusters estimated by the two methods. The DOF parameters estimated by MAP-SVB tend to distribute very broadly, implying that the estimated values may not be so reliable. Moreover, the values of the DOF parameters adopted in MAP-SVB, i.e., the values corresponding to the peak values of the posteriors, tend to be rather small. This results in very heavy tails in Student's t distributions of spike clusters. Therefore, MAP-SVB cannot always be a good method for estimating clustered distributions. In contrast, explicit SVB takes into account the shape of each posterior distribution in the cluster estimation and maintains the size of each distribution in a reasonably narrow range. This means that the confidence level for the estimation is expected to be high. Thus, the estimated clusters tend to show similar shapes and sizes without having extremely heavy tails.


Spike sorting of heterogeneous neuron types by multimodality-weighted PCA and explicit robust variational Bayes.

Takekawa T, Isomura Y, Fukai T - Front Neuroinform (2012)

Typical examples of the posterior distributions of the DOF parameters for the clusters estimated by MAP-SVB and explicit SVB. For MAP-SVB, the vertical lines indicate the values that maximize the distributions. These values are used for the model estimation. For explicit SVB, the vertical lines indicate the average values of the distributions, and explicit SVB takes into account the shapes of the posterior distributions for the model estimation.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3307250&req=5

Figure 7: Typical examples of the posterior distributions of the DOF parameters for the clusters estimated by MAP-SVB and explicit SVB. For MAP-SVB, the vertical lines indicate the values that maximize the distributions. These values are used for the model estimation. For explicit SVB, the vertical lines indicate the average values of the distributions, and explicit SVB takes into account the shapes of the posterior distributions for the model estimation.
Mentions: To explain why explicit SVB is advantageous over MAP-SVB, in Figure 7 we display the posterior distributions of the DOF parameters for the clusters estimated by the two methods. The DOF parameters estimated by MAP-SVB tend to distribute very broadly, implying that the estimated values may not be so reliable. Moreover, the values of the DOF parameters adopted in MAP-SVB, i.e., the values corresponding to the peak values of the posteriors, tend to be rather small. This results in very heavy tails in Student's t distributions of spike clusters. Therefore, MAP-SVB cannot always be a good method for estimating clustered distributions. In contrast, explicit SVB takes into account the shape of each posterior distribution in the cluster estimation and maintains the size of each distribution in a reasonably narrow range. This means that the confidence level for the estimation is expected to be high. Thus, the estimated clusters tend to show similar shapes and sizes without having extremely heavy tails.

Bottom Line: The performance of the proposed method was compared with that of other conventional methods for simulated and experimental data sets.We found that the mPCA efficiently extracts highly informative features as clusters clearly separable in a relatively low-dimensional feature space.Our method showed significantly improved performance in spike sorting of these "difficult" neurons.

View Article: PubMed Central - PubMed

Affiliation: Laboratory for Neural Circuit Theory, RIKEN Brain Science Institute Wako, Japan.

ABSTRACT
This study introduces a new spike sorting method that classifies spike waveforms from multiunit recordings into spike trains of individual neurons. In particular, we develop a method to sort a spike mixture generated by a heterogeneous neural population. Such a spike sorting has a significant practical value, but was previously difficult. The method combines a feature extraction method, which we may term "multimodality-weighted principal component analysis" (mPCA), and a clustering method by variational Bayes for Student's t mixture model (SVB). The performance of the proposed method was compared with that of other conventional methods for simulated and experimental data sets. We found that the mPCA efficiently extracts highly informative features as clusters clearly separable in a relatively low-dimensional feature space. The SVB was implemented explicitly without relying on Maximum-A-Posterior (MAP) inference for the "degree of freedom" parameters. The explicit SVB is faster than the conventional SVB derived with MAP inference and works more reliably over various data sets that include spiking patterns difficult to sort. For instance, spikes of a single bursting neuron may be separated incorrectly into multiple clusters, whereas those of a sparsely firing neuron tend to be merged into clusters for other neurons. Our method showed significantly improved performance in spike sorting of these "difficult" neurons. A parallelized implementation of the proposed algorithm (EToS version 3) is available as open-source code at http://etos.sourceforge.net/.

No MeSH data available.