Recent progress in multi-electrode spike sorting methods

Localization of neurons from extracellular footprints

BACKGROUND

High density microelectrode arrays (HD-MEAs) are now widely used for both in-vitro and in-vivo recordings, as they allow spikes from hundreds of neurons to be recorded simultaneously. Since extracellular recordings do not allow visualization of the recorded neurons, algorithms are needed to estimate their physical positions, especially to track their movements when the are drifting away from recording devices.

NEW METHOD

The objective of this study was to evaluate the performance of multiple algorithms for neuron localization solely from extracellular traces (MEA recordings), either artificial or obtained from mouse retina. The algorithms compared included center-of-mass, monopolar, and grid-based algorithms. The first method is a barycenter calculation. The second algorithm infers the position of the cell using triangulation with the assumption that the neuron behaves as a monopole. Finally, grid-based methods rely on comparing the recorded spike with a projection of spikes of hypothetical neurons with different positions.

RESULTS

The Grid-Based algorithm yielded the most satisfactory outcomes. The center-of-mass exhibited a minimal computational cost, yet its average localization was suboptimal. Monopolar algorithms gave cell localizations with an average error of less than 10μm, but they had considerable variability and a high computational cost. For the grid-based method, the variability was smaller, with satisfactory performance and low computational cost.

COMPARISON WITH EXISTING METHOD(S)

The accuracy of the different localization methods benchmarked in this article had not been properly tested with ground-truth recordings before.

CONCLUSION

The objective of this article is to provide guidance to researchers on the selection of optimal methods for localizing neurons based on MEA recordings.

Comparative analysis of spike-sorters in large-scale brainstem recordings

Recent technological advancements in high-density multi-channel electrodes have made it possible to record large numbers of neurons from previously inaccessible regions. While the performance of automated spike-sorters has been assessed in recordings from cortex, dentate gyrus, and thalamus, the most effective and efficient approach for spike-sorting can depend on the target region due to differing morphological and physiological characteristics. We therefore assessed the performance of five commonly used sorting packages, Kilosort3, MountainSort5, Tridesclous, SpyKING CIRCUS, and IronClust, in recordings from the rostral ventromedial medulla, a region that has been characterized using single-electrode recordings but that is essentially unexplored at the high-density network level. As demonstrated in other brain regions, each sorter produced unique results. Manual curation preferentially eliminated units detected by only one sorter. Kilosort3 and IronClust required the least curation while maintaining the largest number of units, whereas SpyKING CIRCUS and MountainSort5 required substantial curation. Tridesclous consistently identified the smallest number of units. Nonetheless, all sorters successfully identified classically defined RVM physiological cell types. These findings suggest that while the level of manual curation needed may vary across sorters, each can extract meaningful data from this deep brainstem site.

Deep learning-based spike sorting: a survey

Objective.Deep learning is increasingly permeating neuroscience, leading to a rise in signal-processing applications for extracellular recordings. These signals capture the activity of small neuronal populations, necessitating 'spike sorting' to assign action potentials (spikes) to their underlying neurons. With the rise in publications delving into new methodologies and techniques for deep learning-based spike sorting, it is crucial to synthesise these findings critically. This survey provides an in-depth evaluation of the approaches, methodologies and outcomes presented in recent articles, shedding light on the current state-of-the-art.Approach.Twenty-four articles published until December 2023 on deep learning-based spike sorting have been examined. The proposed methods are divided into three sub-problems of spike sorting: spike detection, feature extraction and classification. Moreover, integrated systems, i.e. models that detect spikes and extract features or do classification within a single network, are included.Main results.Although most algorithms have been developed for single-channel recordings, models utilising multi-channel data have already shown promising results, with efficient hardware implementations running quantised models on application-specific integrated circuits and field programmable gate arrays. Convolutional neural networks have been used extensively for spike detection and classification as the data can be processed spatiotemporally while maintaining low-parameter models and increasing generalisation and efficiency. Autoencoders have been mainly utilised for dimensionality reduction, enabling subsequent clustering with standard methods. Also, integrated systems have shown great potential in solving the spike sorting problem from end to end.Significance.This survey explores recent articles on deep learning-based spike sorting and highlights the capabilities of deep neural networks in overcoming associated challenges, but also highlights potential biases of certain models. Serving as a resource for both newcomers and seasoned researchers in the field, this work provides insights into the latest advancements and may inspire future model development.

A Modular Implementation to Handle and Benchmark Drift Correction for High-Density Extracellular Recordings

High-density neural devices are now offering the possibility to record from neuronal populations in vivo at unprecedented scale. However, the mechanical drifts often observed in these recordings are currently a major issue for "spike sorting," an essential analysis step to identify the activity of single neurons from extracellular signals. Although several strategies have been proposed to compensate for such drifts, the lack of proper benchmarks makes it hard to assess the quality and effectiveness of motion correction. In this paper, we present a benchmark study to precisely and quantitatively evaluate the performance of several state-of-the-art motion correction algorithms introduced in the literature. Using simulated recordings with induced drifts, we dissect the origins of the errors performed while applying a motion correction algorithm as a preprocessing step in the spike sorting pipeline. We show how important it is to properly estimate the positions of the neurons from extracellular traces in order to correctly estimate the probe motion, compare several interpolation procedures, and highlight what are the current limits for motion correction approaches.

Decoding the Spike-Band Subthreshold Motor Cortical Activity

Intracortical Brain-Computer Interfaces (iBCI) use single-unit activity (SUA), multiunit activity (MUA) and local field potentials (LFP) to control neuroprosthetic devices. SUA and MUA are usually extracted from the bandpassed recording through amplitude thresholding, while subthreshold data are ignored. Here, we show that subthreshold data can actually be decoded to determine behavioral variables with test set accuracy of up to 100%. Although the utility of SUA, MUA and LFP for decoding behavioral variables has been explored previously, this study investigates the utility of spike-band subthreshold activity exclusively. We provide evidence suggesting that this activity can be used to keep decoding performance at acceptable levels even when SUA quality is reduced over time. To the best of our knowledge, the signals that we derive from the subthreshold activity may be the weakest neural signals that have ever been extracted from extracellular neural recordings, while still being decodable with test set accuracy of up to 100%. These results are relevant for the development of fully data-driven and automated methods for amplitude thresholding spike-band extracellular neural recordings in iBCIs containing thousands of electrodes.

A Fast and Effective Spike Sorting Method Based on Multi-Frequency Composite Waveform Shapes

Accurate spike sorting to the appropriate neuron is crucial for neural activity analysis. To improve spike sorting performance, it is essential to fully leverage each processing step, including filtering, spike detection, feature extraction, and clustering. However, compared to the latter two steps that were widely studied and optimized, the filtering process was largely neglected. In this study, we proposed a fast and effective spike sorting method (MultiFq) based on multi-frequency composite waveform shapes acquired through an optimized filtering process. When combined with the classical PCA-Km spiking sorting algorithm, our proposed MultiFq significantly improved its sorting performance and achieved as high performance as the complex Wave-clus did in both the simulated and in vivo datasets. But, the combined method was about 10 times faster than Wave-clus (0.16 s vs. 2.06 s in simulated datasets; 0.46 s vs. 2.03 s in in vivo datasets). Furthermore, we demonstrated the compatibility of our MultiFq by combining it with other sorting algorithms, which consistently resulted in significant improvement in sorting accuracy with the maximum improvement at 35.04%. The above results demonstrated that our proposed method could significantly improve the sorting performance with low computation cost and good compatibility by leveraging the multi-frequency composite waveform shapes.

Precision of neuronal localization in 2D cell cultures by using high-performance electropolymerized microelectrode arrays correlated with optical imaging

Recently, the development of electronic devices to extracellularly record the simultaneous electrical activities of numerous neurons has been blooming, opening new possibilities to interface and decode neuronal activity. In this work, we tested how the use of EDOT electropolymerization to tune post-fabrication materials could optimize the cell/electrode interface of such devices. Our results showed an improved signal-to-noise ratio, better biocompatibility, and a higher number of neurons detected in comparison with gold electrodes. Then, using such enhanced recordings with 2D neuronal cultures combined with fluorescent optical imaging, we checked the extent to which the positions of the recorded neurons could be estimated solely via their extracellular signatures. Our results showed that assuming neurons behave as monopoles, positions could be estimated with a precision of approximately tens of micrometers.

Spike Sorting of Non-Stationary Data in Successive Intervals Based on Dirichlet Process Mixtures

This paper proposes a new automatic method for spike sorting and tracking non-stationary data based on the Dirichlet Process Mixture (DPM). Data is divided into non-overlapping intervals and mixtures are applied to individual frames rather than to the whole data. In this paper, we have used the information of the previous frame to estimate the cluster parameters of the current interval. Specifically, the means of the clusters in the previous frame are used for estimating the cluster means of the current one, and other parameters are estimated via noninformative priors. The proposed method is capable to track variations in size, shape, or location of clusters as well as detecting the appearance and disappearance of them. We present results in two-dimensional space of first and second principal components (PC1-PC2), but any other feature extraction method leading to the ability of modeling spikes with Normal or t-Student distributions can also be applied. Application of this approach to simulated data and the recordings from anesthetized rat hippocampus confirms its superior performance in comparison to a standard DPM that uses no information from previous frames.

How Do Spike Collisions Affect Spike Sorting Performance?

Recently, a new generation of devices have been developed to record neural activity simultaneously from hundreds of electrodes with a very high spatial density, both for in vitro and in vivo applications. While these advances enable to record from many more cells, they also challenge the already complicated process of spike sorting (i.e., extracting isolated single-neuron activity from extracellular signals). In this work, we used synthetic ground-truth recordings with controlled levels of correlations among neurons to quantitatively benchmark the performance of state-of-the-art spike sorters focusing specifically on spike collisions. Our results show that while modern template-matching-based algorithms are more accurate than density-based approaches, all methods, to some extent, failed to detect synchronous spike events of neurons with similar extracellular signals. Interestingly, the performance of the sorters is not largely affected by the spiking activity in the recordings, with respect to average firing rates and spike-train correlation levels. Since the performances of all modern spike sorting algorithms can be affected as function of the activity of the recorded neurons, scientific claims on correlations and synchrony should be carefully assessed based on the analysis provided in this paper.

Classification of overlapping spikes using convolutional neural networks and long short term memory

Spike sorting is one of the key techniques to understand brain activity. In this paper, we propose a novel deep learning approach based on convolutional neural networks (CNN) and long short term memory (LSTM) to implement overlapping spike sorting. The results of the simulated data demonstrated that the clustering accuracy was greater than 99.9% and 99.0% for non-overlapping spikes and overlapping spikes, respectively. Moreover, the proposed method performed better than our previous deep learning approach named "1D-CNN". In addition, the experimental data recorded from the primary visual cortex of a macaque monkey were used to evaluate the proposed method in a practical application. It was shown that the method could successfully isolate most overlapping spikes of different neurons (ranging from two to five). In summary, the CNN + LSTM method proposed in this paper is of great advantage for overlapping spike sorting with high accuracy. It lays the foundation for application in more challenging works, such as distinguishing the simultaneous recordings of multichannel neuronal activities.

From End to End: Gaining, Sorting, and Employing High-Density Neural Single Unit Recordings

The meaning behind neural single unit activity has constantly been a challenge, so it will persist in the foreseeable future. As one of the most sourced strategies, detecting neural activity in high-resolution neural sensor recordings and then attributing them to their corresponding source neurons correctly, namely the process of spike sorting, has been prevailing so far. Support from ever-improving recording techniques and sophisticated algorithms for extracting worthwhile information and abundance in clustering procedures turned spike sorting into an indispensable tool in electrophysiological analysis. This review attempts to illustrate that in all stages of spike sorting algorithms, the past 5 years innovations' brought about concepts, results, and questions worth sharing with even the non-expert user community. By thoroughly inspecting latest innovations in the field of neural sensors, recording procedures, and various spike sorting strategies, a skeletonization of relevant knowledge lays here, with an initiative to get one step closer to the original objective: deciphering and building in the sense of neural transcript.

Studying Cardiac Neural Network Dynamics: Challenges and Opportunities for Scientific Computing

Neural control of the heart involves continuous modulation of cardiac mechanical and electrical activity to meet the organism's demand for blood flow. The closed-loop control scheme consists of interconnected neural networks with central and peripheral components working cooperatively with each other. These components have evolved to cooperate control of various aspects of cardiac function, which produce measurable "functional" outputs such as heart rate and blood pressure. In this review, we will outline fundamental studies probing the cardiac neural control hierarchy. We will discuss how computational methods can guide improved experimental design and be used to probe how information is processed while closed-loop control is operational. These experimental designs generate large cardio-neural datasets that require sophisticated strategies for signal processing and time series analysis, while presenting the usual large-scale computational challenges surrounding data sharing and reproducibility. These challenges provide unique opportunities for the development and validation of novel techniques to enhance understanding of mechanisms of cardiac pathologies required for clinical implementation.

An Open-Source Wireless Electrophysiological Complex for In Vivo Recording Neuronal Activity in the Rodent's Brain

Multi-electrode arrays (MEAs) are a widely used tool for recording neuronal activity both in vitro/ex vivo and in vivo experiments. In the last decade, researchers have increasingly used MEAs on rodents in vivo. To increase the availability and usability of MEAs, we have created an open-source wireless electrophysiological complex. The complex is scalable, recording the activity of neurons in the brain of rodents during their behavior. Schematic diagrams and a list of necessary components for the fabrication of a wireless electrophysiological complex, consisting of a base charging station and wireless wearable modules, are presented.

Building population models for large-scale neural recordings: Opportunities and pitfalls

Modern recording technologies now enable simultaneous recording from large numbers of neurons. This has driven the development of new statistical models for analyzing and interpreting neural population activity. Here, we provide a broad overview of recent developments in this area. We compare and contrast different approaches, highlight strengths and limitations, and discuss biological and mechanistic insights that these methods provide.

Chronic, Multi-Site Recordings Supported by Two Low-Cost, Stationary Probe Designs Optimized to Capture Either Single Unit or Local Field Potential Activity in Behaving Rats

Rodent models of cognitive behavior have greatly contributed to our understanding of human neuropsychiatric disorders. However, to elucidate the neurobiological underpinnings of such disorders or impairments, animal models are more useful when paired with methods for measuring brain function in awake, behaving animals. Standard tools used for systems-neuroscience level investigations are not optimized for large-scale and high-throughput behavioral battery testing due to various factors including cost, time, poor longevity, and selective targeting limited to measuring only a few brain regions at a time. Here we describe two different "user-friendly" methods for building extracellular electrophysiological probes that can be used to measure either single units or local field potentials in rats performing cognitive tasks. Both probe designs leverage several readily available, yet affordable, commercial products to facilitate ease of production and offer maximum flexibility in terms of brain-target locations that can be scalable (32-64 channels) based on experimental needs. Our approach allows neural activity to be recorded simultaneously with behavior and compared between micro (single unit) and more macro (local field potentials) levels of brain activity in order to gain a better understanding of how local brain regions and their connected networks support cognitive functions in rats. We believe our novel probe designs make collecting electrophysiology data easier and will begin to fill the gap in knowledge between basic and clinical research.

HTsort: Enabling Fast and Accurate Spike Sorting on Multi-Electrode Arrays

Spike sorting is used to classify the spikes (action potentials acquired by physiological electrodes), aiming to identify their respective firing units. Now it has been developed to classify the spikes recorded by multi-electrode arrays (MEAs), with the improvement of micro-electrode technology. However, how to improve classification accuracy and maintain low time complexity simultaneously becomes a difficulty. A fast and accurate spike sorting approach named HTsort is proposed for high-density multi-electrode arrays in this paper. Several improvements have been introduced to the traditional pipeline that is composed of threshold detection and clustering method. First, the divide-and-conquer method is employed to utilize electrode spatial information to achieve pre-clustering. Second, the clustering method HDBSCAN (hierarchical density-based spatial clustering of applications with noise) is used to classify spikes and detect overlapping events (multiple spikes firing simultaneously). Third, the template merging method is used to merge redundant exported templates according to the template similarity and the spatial distribution of electrodes. Finally, the template matching method is used to resolve overlapping events. Our approach is validated on simulation data constructed by ourselves and publicly available data and compared to other state-of-the-art spike sorters. We found that the proposed HTsort has a more favorable trade-off between accuracy and time consumption. Compared with MountainSort and SpykingCircus, the time consumption is reduced by at least 40% when the number of electrodes is 64 and below. Compared with HerdingSpikes, the classification accuracy can typically improve by more than 10%. Meanwhile, HTsort exhibits stronger robustness against background noise than other sorters. Our more sophisticated spike sorter would facilitate neurophysiologists to complete spike sorting more quickly and accurately.

A Unified Optimization Model of Feature Extraction and Clustering for Spike Sorting

Spike sorting technologies support neuroscientists to access the neural activity with single-neuron or single-action-potential resolutions. However, conventional spike sorting technologies perform the feature extraction and the clustering separately after the spikes are well detected. It not only induces many redundant processes, but it also yields a lower accuracy and an unstable result especially when noises and/or overlapping spikes exist in the dataset. To address these issues, this paper proposes a unified optimization model integrating the feature extraction and the clustering for spike sorting. Unlike the widely used combination strategies, i.e., performing the principal component analysis (PCA) for spike feature extraction and the K-means (KM) for clustering in sequence, interestingly, this paper finds the solution of the proposed unified model by iteratively performing PCA and KM-like procedures. Subsequently, by embedding the K-means++ strategy in KM-like initializing and a comparison updating rule in the solving process, the proposed model can well handle the noises and overlapping interference as well as enjoy a high accuracy and a low computational complexity. Finally, an automatic spike sorting method is derived after taking the best of the clustering validity indices into the proposed model. The extensive numerical simulation results on both synthetic and real-world datasets confirm that our proposed method outperforms the related state-of-the-art approaches.

MEArec: A Fast and Customizable Testbench Simulator for Ground-truth Extracellular Spiking Activity

When recording neural activity from extracellular electrodes, both in vivo and in vitro, spike sorting is a required and very important processing step that allows for identification of single neurons’ activity. Spike sorting is a complex algorithmic procedure, and in recent years many groups have attempted to tackle this problem, resulting in numerous methods and software packages. However, validation of spike sorting techniques is complicated. It is an inherently unsupervised problem and it is hard to find universal metrics to evaluate performance. Simultaneous recordings that combine extracellular and patch-clamp or juxtacellular techniques can provide ground-truth data to evaluate spike sorting methods. However, their utility is limited by the fact that only a few cells can be measured at the same time. Simulated ground-truth recordings can provide a powerful alternative mean to rank the performance of spike sorters. We present here MEArec, a Python-based software which permits flexible and fast simulation of extracellular recordings. MEArec allows users to generate extracellular signals on various customizable electrode designs and can replicate various problematic aspects for spike sorting, such as bursting, spatio-temporal overlapping events, and drifts. We expect MEArec will provide a common testbench for spike sorting development and evaluation, in which spike sorting developers can rapidly generate and evaluate the performance of their algorithms.

A facile and comprehensive algorithm for electrical response identification in mouse retinal ganglion cells

Retinal prostheses can restore the basic visual function of patients with retinal degeneration, which relies on effective electrical stimulation to evoke the physiological activities of retinal ganglion cells (RGCs). Current electrical stimulation strategies have defects such as unstable effects and insufficient stimulation positions, therefore, it is crucial to determine the optimal pulse parameters for precise and safe electrical stimulation. Biphasic voltages (cathode-first) with a pulse width of 25 ms and different amplitudes were used to ex vivo stimulate RGCs of three wild-type (WT) mice using a commercial microelectrode array (MEA) recording system. An algorithm is developed to automatically realize both spike-sorting and electrical response identification for the spike signals recorded. Measured from three WT mouse retinas, the total numbers of RGC units and responsive RGC units were 1193 and 151, respectively. In addition, the optimal pulse amplitude range for electrical stimulation was determined to be 0.43 V-1.3 V. The processing results of the automatic algorithm we proposed shows high consistency with those using traditional manual processing. We anticipate the new algorithm can not only speed up the elaborate electrophysiological data processing, but also optimize pulse parameters for the electrical stimulation strategy of neural prostheses.

Electrophysiological Analysis of Brain Organoids: Current Approaches and Advancements

Brain organoids, or cerebral organoids, have become widely used to study the human brain in vitro. As pluripotent stem cell-derived structures capable of self-organization and recapitulation of physiological cell types and architecture, brain organoids bridge the gap between relatively simple two-dimensional human cell cultures and non-human animal models. This allows for high complexity and physiological relevance in a controlled in vitro setting, opening the door for a variety of applications including development and disease modeling and high-throughput screening. While technologies such as single cell sequencing have led to significant advances in brain organoid characterization and understanding, improved functional analysis (especially electrophysiology) is needed to realize the full potential of brain organoids. In this review, we highlight key technologies for brain organoid development and characterization, then discuss current electrophysiological methods for brain organoid analysis. While electrophysiological approaches have improved rapidly for two-dimensional cultures, only in the past several years have advances been made to overcome limitations posed by the three-dimensionality of brain organoids. Here, we review major advances in electrophysiological technologies and analytical methods with a focus on advances with applicability for brain organoid analysis.

An Accurate and Robust Method for Spike Sorting Based on Convolutional Neural Networks

In the fields of neuroscience and biomedical signal processing, spike sorting is a crucial step to extract the information of single neurons from extracellular recordings. In this paper, we propose a novel deep learning approach based on one-dimensional convolutional neural networks (1D-CNNs) to implement accurate and robust spike sorting. The results of the simulated data demonstrated that the clustering accuracy in most datasets was greater than 99%, despite the multiple levels of noise and various degrees of overlapped spikes. Moreover, the proposed method performed significantly better than the state-of-the-art method named “WMsorting” and a deep-learning-based multilayer perceptron (MLP) model. In addition, the experimental data recorded from the primary visual cortex of a macaque monkey were used to evaluate the proposed method in a practical application. It was shown that the method could successfully isolate most spikes of different neurons (ranging from two to five) by training the 1D-CNN model with a small number of manually labeled spikes. Considering the above, the deep learning method proposed in this paper is of great advantage for spike sorting with high accuracy and strong robustness. It lays the foundation for application in more challenging works, such as distinguishing overlapped spikes and the simultaneous sorting of multichannel recordings.

A stacked sparse auto-encoder and back propagation network model for sensory event detection via a flexible ECoG

Current prostheses are limited in their ability to provide direct sensory feedback to users with missing limb. Several efforts have been made to restore tactile sensation to amputees but the somatotopic tactile feedback often results in unnatural sensations, and it is yet unclear how and what information the somatosensory system receives during voluntary movement. The present study proposes an efficient model of stacked sparse autoencoder and back propagation neural network for detecting sensory events from a highly flexible electrocorticography (ECoG) electrode. During the mechanical stimulation with Von Frey (VF) filament on the plantar surface of rats' foot, simultaneous recordings of tactile afferent signals were obtained from primary somatosensory cortex (S1) in the brain. In order to achieve a model with optimal performance, Particle Swarm Optimization and Adaptive Moment Estimation (Adam) were adopted to select the appropriate number of neurons, hidden layers and learning rate of each sparse auto-encoder. We evaluated the stimulus-evoked sensation by using an automated up-down (UD) method otherwise called UDReader. The assessment of tactile thresholds with VF shows that the right side of the hind-paw was significantly more sensitive at the tibia-(p = 6.50 × 10^-4), followed by the saphenous-(p = 7.84 × 10^-4), and sural-(p = 8.24 × 10^-4). We then validated our proposed model by comparing with the state-of-the-art methods, and recorded accuracy of 98.8%, sensitivity of 96.8%, and specificity of 99.1%. Hence, we demonstrated the effectiveness of our algorithms in detecting sensory events through flexible ECoG recordings which could be a viable option in restoring somatosensory feedback.

SpikeForest, reproducible web-facing ground-truth validation of automated neural spike sorters

Spike sorting is a crucial step in electrophysiological studies of neuronal activity. While many spike sorting packages are available, there is little consensus about which are most accurate under different experimental conditions. SpikeForest is an open-source and reproducible software suite that benchmarks the performance of automated spike sorting algorithms across an extensive, curated database of ground-truth electrophysiological recordings, displaying results interactively on a continuously-updating website. With contributions from eleven laboratories, our database currently comprises 650 recordings (1.3 TB total size) with around 35,000 ground-truth units. These data include paired intracellular/extracellular recordings and state-of-the-art simulated recordings. Ten of the most popular spike sorting codes are wrapped in a Python package and evaluated on a compute cluster using an automated pipeline. SpikeForest documents community progress in automated spike sorting, and guides neuroscientists to an optimal choice of sorter and parameters for a wide range of probes and brain regions.

Characterization and Classification of Electrophysiological Signals Represented as Visibility Graphs Using the Maxclique Graph

Detection, characterization and classification of patterns within time series from electrophysiological signals have been a challenge for neuroscientists due to their complexity and variability. Here, we aimed to use graph theory to characterize and classify waveforms within biological signals using maxcliques as a feature for a deep learning method. We implemented a compact and easy to visualize algorithm and interface in Python. This software uses time series as input. We applied the maxclique graph operator in order to obtain further graph parameters. We extracted features of the time series by processing all graph parameters through K-means, one of the simplest unsupervised machine learning algorithms. As proof of principle, we analyzed integrated electrical activity of XII nerve to identify waveforms. Our results show that the use of maxcliques allows identification of two distinct types of waveforms that match expert classification. We propose that our method can be a useful tool to characterize and classify other electrophysiological signals in a short time and objectively. Reducing the classification time improves efficiency for further analysis in order to compare between treatments or conditions, e.g., pharmacological trials, injuries, or neurodegenerative diseases.

Independent Component Analysis for Fully Automated Multi-Electrode Array Spike Sorting

In neural electrophysiology, spike sorting allows to separate different neurons from extracellularly measured recordings. It is an essential processing step in order to understand neural activity and it is an unsupervised problem in nature, since no ground truth information is available. There are several available spike sorting packages, but many of them require a manual intervention to curate the results, which makes the process time consuming and hard to reproduce. Here, we focus on high-density Multi-Electrode Array (MEA) recordings and we present a fully automated pipeline based on Independent Component Analysis (ICA). While ICA has been previously investigated for spike sorting, it has never been compared with fully automated state-of-the-art algorithms. We use realistic simulated datasets to compare the spike sorting performance in terms of complexity, signal-to-noise ratio, and recording duration. We show that an ICA-based fully automated spike sorting approach can be a viable alternative approach due to its precision and robustness, but it needs to be optimized for time constraints and requires sufficient density of electrodes to cover active neurons in the proximity of the MEA.

Human motor decoding from neural signals: a review

Many people suffer from movement disability due to amputation or neurological diseases. Fortunately, with modern neurotechnology now it is possible to intercept motor control signals at various points along the neural transduction pathway and use that to drive external devices for communication or control. Here we will review the latest developments in human motor decoding. We reviewed the various strategies to decode motor intention from human and their respective advantages and challenges. Neural control signals can be intercepted at various points in the neural signal transduction pathway, including the brain (electroencephalography, electrocorticography, intracortical recordings), the nerves (peripheral nerve recordings) and the muscles (electromyography). We systematically discussed the sites of signal acquisition, available neural features, signal processing techniques and decoding algorithms in each of these potential interception points. Examples of applications and the current state-of-the-art performance were also reviewed. Although great strides have been made in human motor decoding, we are still far away from achieving naturalistic and dexterous control like our native limbs. Concerted efforts from material scientists, electrical engineers, and healthcare professionals are needed to further advance the field and make the technology widely available in clinical use.

A normalized template matching method for improving spike detection in extracellular voltage recordings

Spike sorting is the process of detecting and clustering action potential waveforms of putative single neurons from extracellular voltage recordings. Typically, spike detection uses a fixed voltage threshold and shadow period, but this approach often misses spikes during high firing rate epochs or noisy conditions. We developed a simple, data-driven spike detection method using a scaled form of template matching, based on the sliding cosine similarity between the extracellular voltage signal and mean spike waveforms of candidate single units. Performance was tested in whisker somatosensory cortex (S1) of anesthetized mice in vivo. The method consistently detected whisker-evoked spikes that were missed by the standard fixed threshold. Detection was improved most for spikes evoked by strong stimuli (40–70% increase), improved less for weaker stimuli, and unchanged for spontaneous spiking. This represents improved detection during spatiotemporally dense spiking, and yielded sharper sensory tuning estimates. We also benchmarked performance using computationally generated voltage data. Template matching detected ~85–90% of spikes compared to ~70% for the standard fixed threshold method, and was more tolerant to high firing rates and simulated recording noise. Thus, a simple template matching approach substantially improves detection of single-unit spiking for cortical physiology.

Evaluation of Spike Sorting Algorithms: Application to Human Subthalamic Nucleus Recordings and Simulations

An important prerequisite for the analysis of spike synchrony in extracellular recordings is the extraction of single-unit activity from the multi-unit signal. To identify single units, potential spikes are separated with respect to their potential neuronal origins ('spike sorting'). However, different sorting algorithms yield inconsistent unit assignments, which seriously influences subsequent spike train analyses. We aim to identify the best sorting algorithm for subthalamic nucleus recordings of patients with Parkinson's disease (experimental data ED). Therefore, we apply various prevalent algorithms offered by the 'Plexon Offline Sorter' and evaluate the sorting results. Since this evaluation leaves us unsure about the best algorithm, we apply all methods again to artificial data (AD) with known ground truth. AD consists of pairs of single units with different shape similarity embedded in the background noise of the ED. The sorting evaluation depicts a significant influence of the respective methods on the single unit assignments. We find a high variability in the sortings obtained by different algorithms that increases with single units shape similarity. We also find significant differences in the resulting firing characteristics. We conclude that Valley-Seeking algorithms produce the most accurate result if the exclusion of artifacts as unsorted events is important. If the latter is less important ('clean' data) the K-Means algorithm is a better option. Our results strongly argue for the need of standardized validation procedures based on ground truth data. The recipe suggested here is simple enough to become a standard procedure.

Electrophysiological monitoring of inhibition in mammalian species, from rodents to humans

GABAergic interneurons constitute a highly diverse family of neurons that play a critical role in cortical functions. Due to their prominent role in cortical network dynamics, genetic, developmental, or other dysfunctions in GABAergic neurons have been linked to neurological disorders such as epilepsy. Thus, it is crucial to investigate the interaction of these various neurons and to develop methods to specifically and directly monitor inhibitory activity in vivo. While research in small mammals has benefited from a wealth of recent technological development, bridging the gap to large mammals and humans remains a challenge. This is of particular interest since single neuron monitoring with intracranial electrodes in epileptic patients is developing quickly, opening new avenues for understanding the role of different cell types in epilepsy. Here, we review currently available techniques that monitor inhibitory activity in the brain and the respective validations in rodents. Finally, we discuss the future developments of these techniques and how knowledge from animal research can be translated to the study of neuronal circuit dynamics in the human brain.

Continuing progress of spike sorting in the era of big data

Engineering efforts are currently attempting to build devices capable of collecting neural activity from one million neurons in the brain. Part of this effort focuses on developing dense multiple-electrode arrays, which require post-processing via 'spike sorting' to extract neural spike trains from the raw signal. Gathering information at this scale will facilitate fascinating science, but these dreams are only realizable if the spike sorting procedure and data pipeline are computationally scalable, at or superior to hand processing, and scientifically reproducible. These challenges are all being amplified as the data scale continues to increase. In this review, recent efforts to attack these challenges are discussed, which have primarily focused on increasing accuracy and reliability while being computationally scalable. These goals are addressed by adding additional stages to the data processing pipeline and using divide-and-conquer algorithmic approaches. These recent developments should prove useful to most research groups regardless of data scale, not just for cutting-edge devices.

Spike sorting with Gaussian mixture models

The shape of extracellularly recorded action potentials is a product of several variables, such as the biophysical and anatomical properties of the neuron and the relative position of the electrode. This allows isolating spikes of different neurons recorded in the same channel into clusters based on waveform features. However, correctly classifying spike waveforms into their underlying neuronal sources remains a challenge. This process, called spike sorting, typically consists of two steps: (1) extracting relevant waveform features (e.g., height, width), and (2) clustering them into non-overlapping groups believed to correspond to different neurons. In this study, we explored the performance of Gaussian mixture models (GMMs) in these two steps. We extracted relevant features using a combination of common techniques (e.g., principal components, wavelets) and GMM fitting parameters (e.g., Gaussian distances). Then, we developed an approach to perform unsupervised clustering using GMMs, estimating cluster properties in a data-driven way. We found the proposed GMM-based framework outperforms previously established methods in simulated and real extracellular recordings. We also discuss potentially better techniques for feature extraction than the widely used principal components. Finally, we provide a friendly graphical user interface to run our algorithm, which allows manual adjustments.

Reconstruction of Functional Connectivity from Multielectrode Recordings and Calcium Imaging

In the last two decades, increasing research efforts in neuroscience have been focused on determining both structural and functional connectivity of brain circuits, with the main goal of relating the wiring diagram of neuronal systems to their emerging properties, from the microscale to the macroscale. While combining multisite parallel recordings with structural circuits' reconstruction in vivo is still very challenging, the reductionist in vitro approach based on neuronal cultures offers lower technical difficulties and is much more stable under control conditions. In this chapter, we present different approaches to infer the connectivity of cultured neuronal networks using multielectrode array or calcium imaging recordings. We first formally introduce the used methods, and then we will describe into details how those methods were applied in case studies. Since multielectrode array and calcium imaging recordings provide distinct and complementary spatiotemporal features of neuronal activity, in this chapter we present the strategies implemented with the two different methodologies in distinct sections.

Spike detection and spike sorting with a hidden Markov model improves offline decoding of motor cortical recordings

OBJECTIVE

Detection and sorting (classification) of action potentials from extracellular recordings are two important pre-processing steps for brain-computer interfaces (BCIs) and some neuroscientific studies. Traditional approaches perform these two steps serially, but using shapes of action potential waveforms during detection, i.e. combining the two steps, may lead to better performance, especially during high noise. We propose a hidden Markov model (HMM) based method for combined detecting and sorting of spikes, with the aim of improving the final decoding accuracy of BCIs.

APPROACH

The states of the HMM indicate whether there is a spike, what unit a spike belongs to, and the time course within a waveform. The HMM outputs probabilities of spike detection, and from this we can calculate expectations of spike counts in time bins, which can replace integer spike counts as input to BCI decoders. We evaluate the HMM method on simulated spiking data. We then examine the impact of using this method on decoding real neural data recorded from primary motor cortex of two Rhesus monkeys.

MAIN RESULTS

Our comparisons on simulated data to detection-then-sorting approaches and combined detection-and-sorting algorithms indicate that the HMM method performs more accurately at detection and sorting (0.93 versus 0.73 spike count correlation, 0.73 versus 0.49 adjusted mutual information). On real neural data, the HMM method led to higher adjusted mutual information between spike counts and kinematics (monkey K: 0.034 versus 0.027; monkey M: 0.033 versus 0.022) and better neuron encoding model predictions (K: 0.016 dB improvement; M: 0.056 dB improvement). Lastly, the HMM method facilitated higher offline decoding accuracy (Kalman filter, K: 8.5% mean squared error reduction, M: 18.6% reduction).

SIGNIFICANCE

The HMM spike detection and sorting method offers a new approach to spike pre-processing for BCIs and neuroscientific studies.

Spike sorting based on shape, phase, and distribution features, and K-TOPS clustering with validity and error indices

Spike sorting is one of the most important data analysis problems in neurophysiology. The precision in all steps of the spike-sorting procedure critically affects the accuracy of all subsequent analyses. After data preprocessing and spike detection have been carried out properly, both feature extraction and spike clustering are the most critical subsequent steps of the spike-sorting procedure. The proposed spike sorting approach comprised a new feature extraction method based on shape, phase, and distribution features of each spike (hereinafter SS-SPDF method), which reveal significant information of the neural events under study. In addition, we applied an efficient clustering algorithm based on K-means and template optimization in phase space (hereinafter K-TOPS) that included two integrative clustering measures (validity and error indices) to verify the cohesion-dispersion among spike events during classification and the misclassification of clustering, respectively. The proposed method/algorithm was tested on both simulated data and real neural recordings. The results obtained for these datasets suggest that our spike sorting approach provides an efficient way for sorting both single-unit spikes and overlapping waveforms. By analyzing raw extracellular recordings collected from the rostral-medial prefrontal cortex (rmPFC) of behaving rabbits during classical eyeblink conditioning, we have demonstrated that the present method/algorithm performs better at classifying spikes and neurons and at assessing their modulating properties than other methods currently used in neurophysiology.

Information Entropy Production of Maximum Entropy Markov Chains from Spike Trains

The spiking activity of neuronal networks follows laws that are not time-reversal symmetric; the notion of pre-synaptic and post-synaptic neurons, stimulus correlations and noise correlations have a clear time order. Therefore, a biologically realistic statistical model for the spiking activity should be able to capture some degree of time irreversibility. We use the thermodynamic formalism to build a framework in the context maximum entropy models to quantify the degree of time irreversibility, providing an explicit formula for the information entropy production of the inferred maximum entropy Markov chain. We provide examples to illustrate our results and discuss the importance of time irreversibility for modeling the spike train statistics.

Sorting Overlapping Spike Waveforms from Electrode and Tetrode Recordings

One of the outstanding problems in the sorting of neuronal spike trains is the resolution of overlapping spikes. Resolving these spikes can significantly improve a range of analyses, such as response variability, correlation, and latency. In this paper, we describe a partially automated method that is capable of resolving overlapping spikes. After constructing template waveforms for well-isolated and distinct single units, we generated pair-wise combinations of those templates at all possible time shifts from each other. Subsequently, overlapping waveforms were identified by cluster analysis, and then assigned to their respective single-unit combinations. We examined the performance of this method using simulated data from an earlier study, and found that we were able to resolve an average of 83% of the overlapping waveforms across various signal-to-noise ratios, an improvement of approximately 32% over the results reported in the earlier study. When applied to additional simulated data sets generated from single-electrode and tetrode recordings, we were able to resolve 91% of the overlapping waveforms with a false positive rate of 0.19% for single-electrode data, and 95% of the overlapping waveforms with a false positive rate of 0.27% for tetrode data. We also applied our method to electrode and tetrode data recorded from the primary visual cortex, and the results obtained for these datasets suggest that our method provides an efficient means of sorting overlapping waveforms. This method can easily be added as an extra step to commonly used spike sorting methods, such as KlustaKwik and MClust software packages, and can be applied to datasets that have already been sorted using these methods.