Spike sorting is a fundamental step in extracting single-unit activity from neural ensemble recordings, which play an important role in basic neuroscience and neurotechnologies. A few algorithms have been applied in spike sorting. However, when noise level or waveform similarity becomes relatively high, their robustness still faces a big challenge. In this study, we propose a spike sorting method combining Linear Discriminant Analysis (LDA) and Density Peaks (DP) for feature extraction and clustering. Relying on the joint optimization of LDA and DP: DP provides more accurate classification labels for LDA, LDA extracts more discriminative features to cluster for DP, and the algorithm achieves high performance after iteration. We first compared the proposed LDA-DP algorithm with several algorithms on one publicly available simulated dataset and one real rodent neural dataset with different noise levels. We further demonstrated the performance of the LDA-DP method on a real neural dataset from non-human primates with more complex distribution characteristics. The results show that our LDA-DP algorithm extracts a more discriminative feature subspace and achieves better cluster quality than previously established methods in both simulated and real data. Especially in the neural recordings with high noise levels or waveform similarity, the LDA-DP still yields a robust performance with automatic detection of the number of clusters. The proposed LDA-DP algorithm achieved high sorting accuracy and robustness to noise, which offers a promising tool for spike sorting and facilitates the following analysis of neural population activity.

One of the most concerned problems in neuroscience is how neurons communicate and convey information through spikes. There is abundant evidence in sensory systems to support the use of precise timing of spikes to encode information. However, it remains unknown whether precise temporal patterns could be generated to drive output in the primary motor cortex (M1), a brain area containing ample recurrent connections that may destroy temporal fidelity. Here, we used a novel brain-machine interface that mapped the temporal order and precision of motor cortex activity to the auditory cursor and reward to guide the generation of precise temporal patterns in M1. During the course of learning, rats performed the “temporal neuroprosthetics” in a goal-directed manner with increasing proficiency. Precisely timed spiking activity in M1 was volitionally and robustly produced under this “temporal neuroprosthetics”, demonstrating the feasibility of M1 implementing temporal codes. Population analysis showed that the local network was coordinated in a fine time scale as the overall excitation heightened. Furthermore, we found that the directed connection between neurons assigned to directly control the output (“direct neurons”) was strengthened throughout learning, as well as connections in the subnetwork that contains direct neurons. Network models revealed that excitatory gain and strengthening of subnetwork connectivity transitioned neural states to a more synchronous regime, which improved the sensitivity for coincidence detection and, thus, the precision of spike patterns. Therefore, our results suggested the recurrent connections facilitate the implementation of precise temporal patterns instead of impairing them, which provided new perspectives on the fine-timescale activity and dynamics of M1.

No abstract

No abstract