Spike sorting is a crucial step to extract information from extracellular recordings. With new recording opportunities provided by the development of new electrodes that allow monitoring hundreds of neurons simultaneously, the scenario for the new generation of algorithms is both exciting and challenging. However, this will require a new approach to the problem and the development of a common reference framework to quickly assess the performance of new algorithms. In this work, we review the basic concepts of spike sorting, including the requirements for different applications, together with the problems faced by presently available algorithms. We conclude by proposing a roadmap stressing the crucial points to be addressed to support the neuroscientific research of the near future.

No abstract

The most widely used spike-sorting algorithms are semiautomatic in practice, requiring manual tuning of the automatic solution to achieve good performance. In this work, we propose a new fully automatic spike-sorting algorithm that can capture multiple clusters of different sizes and densities. In addition, we introduce an improved feature selection method, by using a variable number of wavelet coefficients, based on the degree of non-Gaussianity of their distributions. We evaluated the performance of the proposed algorithm with real and simulated data. With real data from single-channel recordings, in ~95% of the cases the new algorithm replicated, in an unsupervised way, the solutions obtained by expert sorters, who manually optimized the solution of a previous semiautomatic algorithm. This was done while maintaining a low number of false positives. With simulated data from single-channel and tetrode recordings, the new algorithm was able to correctly detect many more neurons compared with previous implementations and also compared with recently introduced algorithms, while significantly reducing the number of false positives. In addition, the proposed algorithm showed good performance when tested with real tetrode recordings.NEW & NOTEWORTHY We propose a new fully automatic spike-sorting algorithm, including several steps that allow the selection of multiple clusters of different sizes and densities. Moreover, it defines the dimensionality of the feature space in an unsupervised way. We evaluated the performance of the algorithm with real and simulated data, from both single-channel and tetrode recordings. The proposed algorithm was able to outperform manual sorting from experts and other recent unsupervised algorithms.

Summary The relationship between the firing of single cells and local field potentials (LFPs) has received increasing attention, with studies in animals [1-11] and humans [12-14]. Recordings in the human medial temporal lobe (MTL) have demonstrated the existence of neurons with selective and invariant responses [15], with a relatively late but precise response onset around 300 ms after stimulus presentation [16-18] and firing only upon conscious recognition of the stimulus [19]. This represents a much later onset than expected from direct projections from inferotemporal cortex [16, 18]. The neural mechanisms underlying this onset remain unclear. To address this issue, we performed a joint analysis of single-cell and LFP responses during a visual recognition task. Single-neuron responses were preceded by a global LFP deflection in the theta range. In addition, there was a local and stimulus-specific increase in the single-trial gamma power. These LFP responses correlated with conscious recognition. The timing of the neurons’ firing was phase locked to these LFP responses. We propose that whereas the gamma phase locking reflects the activation of local networks encoding particular recognized stimuli, the theta phase locking reflects a global activation that provides a temporal window for processing consciously perceived stimuli in the MTL.

Recordings from individual neurons in patients who are implanted with depth electrodes for clinical reasons have opened the possibility to narrow down the gap between neurophysiological studies in animals and non-invasive (e.g. functional magnetic resonance imaging, electroencephalogram, magnetoencephalography) investigations in humans. Here we provide a description of the main procedures for electrode implantation and recordings, the experimental paradigms used and the main steps for processing the data. We also present key characteristics of the so-called ‘concept cells’, neurons in the human medial temporal lobe with selective and invariant responses that represent the meaning of the stimulus, and discuss their proposed role in declarative memory. Finally, we present novel results dealing with the stability of the representation given by these neurons, by studying the effect of stimulus repetition in the strength of the responses. In particular, we show that, after an initial decay, the response strength reaches an asymptotic value after approximately 15 presentations that remains above baseline for the whole duration of the experiment.