Spike sorting with Kilosort4

Nature Methods, Journal Year: 2024, Volume and Issue: 21(5), P. 914 - 921

Published: April 8, 2024

Abstract Spike sorting is the computational process of extracting firing times single neurons from recordings local electrical fields. This an important but hard problem in neuroscience, made complicated by nonstationarity and dense overlap fields between nearby neurons. To address spike-sorting problem, we have been openly developing Kilosort framework. Here describe various algorithmic steps introduced different versions Kilosort. We also report development Kilosort4, a version with substantially improved performance due to clustering algorithms inspired graph-based approaches. test Kilosort, developed realistic simulation framework that uses densely sampled real experiments generate nonstationary spike waveforms noise. found nearly all outperformed other on variety simulated conditions Kilosort4 performed best cases, correctly identifying even low amplitudes small spatial extents high drift conditions.

Language: Английский

---

DREDge: robust motion correction for high-density extracellular recordings across species

Nature Methods, Journal Year: 2025, Volume and Issue: unknown

Published: March 6, 2025

Language: Английский

---

PhysMAP - interpretablein vivoneuronal cell type identification using multi-modal analysis of electrophysiological data

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Feb. 28, 2024

Abstract Cells of different types perform diverse computations and coordinate their activity during sensation, perception, action. While electrophysiological approaches can measure the many neurons simultaneously, assigning cell type labels to these is an open problem. Here, we develop PhysMAP, a framework that weighs multiple modalities simultaneously in unsupervised manner obtain interpretable representation separates by type. PhysMAP superior any single modality identifying neuronal such as excitatory pyramidal, PV + interneurons, SOM interneurons with high confidence both juxtacellular extracellular recordings from areas mouse brain. built on ground truth data be used for classifying new existing datasets, thus facilitate simultaneous assessment coordinated dynamics behavior.

Language: Английский

---

Neural Activity in Quarks Language: Lattice Field Theory for a Network of Real Neurons

Entropy, Journal Year: 2024, Volume and Issue: 26(6), P. 495 - 495

Published: June 6, 2024

Brain–computer interfaces have seen extraordinary surges in developments recent years, and a significant discrepancy now exists between the abundance of available data limited headway made achieving unified theoretical framework. This becomes particularly pronounced when examining collective neural activity at micro meso scale, where coherent formalization that adequately describes interactions is still lacking. Here, we introduce mathematical framework to analyze systems natural neurons interpret related empirical observations terms lattice field theory, an established paradigm from particle physics statistical mechanics. Our methods are tailored chronic interfaces, especially spike rasters measurements single neuron activity, generalize maximum entropy model for networks so time evolution system also taken into account. obtained by bridging neuroscience, paving way physics-inspired models neocortex.

Language: Английский

---

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

eNeuro, Journal Year: 2024, Volume and Issue: 11(2), P. ENEURO.0229 - 23.2023

Published: Jan. 18, 2024

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

Language: Английский

---

Integrating multimodal data to understand cortical circuit architecture and function

Nature Neuroscience, Journal Year: 2025, Volume and Issue: unknown

Published: March 24, 2025

Language: Английский

---

NeuroCarto: A Toolkit for Building Custom Read-out Channel Maps for High Electrode-count Neural Probes

Neuroinformatics, Journal Year: 2025, Volume and Issue: 23(1), P. 1 - 16

Published: Jan. 4, 2025

Neuropixels probes contain thousands of electrodes across one or more shanks and are sufficiently small to allow chronic recording neural activity in freely behaving animals. However, the joint increase number miniaturization probe package has led a compromise which groups share single read-out channel only fraction can be read out at any given time. Experimenters then face challenge selecting subset (i.e., map) that both covers brain regions interest adheres restrictions underlying hardware. Here, we present NeuroCarto, Python toolkit GUI simplify construction custom map for probes. We describe general iterative approach select provide specific implementation allows experimenters specify blueprint along desired local electrode density. NeuroCarto assists generating from visualizes potential conflicts. showcase utility an experimental workflow simultaneously record dorsal ventral hippocampus with 4-shank 2.0 moving mice.

Language: Английский

---

Large-scale single-neuron recording using the uFINE array in human cortex

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: April 19, 2025

SUMMARY Monitoring neural population activity at single-cell resolution is essential for driving fundamental research and clinical innovations. However, translating advanced recording techniques from animal models to humans remains a significant challenge. Flexible electrodes have recently emerged as powerful tools large-scale single-unit recordings due their superior biocompatibility high density. Here, we demonstrate reliable, high-density during intraoperative procedures in human patients using u ltra- F lexible I mplantable N eural E lectrode (uFINE) arrays. The uFINE array exhibited sufficient mechanical robustness maintain structural integrity throughout surgical operations. We successfully recorded 616 single units 10 patients, with up 135 simultaneously recorded. flexibility of minimized signal disturbances brain pulsations, enabling stable continuous detection. Stimulus response tuning were observed the level individual neurons awake patients. This uFINE-based approach offers unique opportunities investigate human-specific cognitive functions develop next-generation brain-machine interfaces.

Language: Английский

---

Optimal inter-electrode distances for maximizing single unit yield per electrode in neural recordings

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: April 17, 2025

ABSTRACT State-of-the-art high-density multielectrode arrays enable the recording of simultaneous spiking activity from hundreds neurons. Although significant efforts have been dedicated to enhancing neural devices and developing more efficient sorting algorithms, there has relatively less focus on allocation microelectrodes—a factor that undeniably affects spike effectiveness ultimately total number detected Here, we systematically examined relationship between optimal electrode spacing efficiency by creating virtual sparser layouts recordings through spatial downsampling. We assessed performance comparing quantity well-isolated single units per in sparse configurations across various brain regions (neocortex thalamus) species (rat, mouse, human). Enabling theoretical estimation arrangements, complement experimental results with a geometrical modeling framework. Contrary general assumption higher density inherently leads sorting, both our reveal clear optimum for specific regions. demonstrate carefully choosing distances could yield 1.7 3 times increase efficiency. These findings emphasize necessity species- region-specific microelectrode design optimization.

Language: Английский

---

A simple, inexpensive battery-powered homeothermic warming pad for mice and rats

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2025, Volume and Issue: unknown

Published: May 1, 2025

Abstract Background Anesthesia decreases core body temperature, and this seriously compromises the physiological status of an experimental animal. Hypothermia alters many aspects neural function. When recording nervous system activity in anesthetized animals, their temperature must be stabilized. New method This report describes inexpensive, battery-powered, temperature-controlled warming pad for mice rats documents its validity utility. The device is portable, making it convenient researchers who conduct procedures such as surgical preparations one location transport animal to another recordings. Results keeps normothermic ±0.7° over 6 hours without supplemental warmth (e.g., heat lamp), despite >15° differential between ambient room temperature. We demonstrate how can used vivo imaging neuronal a prolonged period mice. Comparison with existing methods Commercial heating pads small animals are expensive, somewhat bulky, require power cords 120/240V source. Transporting from (e.g. suite) rig) involves moving cords. Moreover, commercial devices not always compatible custom stereotaxic frames, microscope stages, or holding boxes. described here small, readily adaptable set ups. Conclusion homeothermic provides simple maintaining animals. It constructed under 30 minutes, components available, cost less than $100. exceptionally useful experiments on rats. Highlights Temperature-controlled rodents Low cost, well $100 Components available online suppliers Can assembled minutes Does specialty tools stable robust

Language: Английский

---