Lattice physics approaches for neural networks

iScience, Journal Year: 2024, Volume and Issue: 27(12), P. 111390 - 111390

Published: Nov. 15, 2024

Modern neuroscience has evolved into a frontier field that draws on numerous disciplines, resulting in the flourishing of novel conceptual frames primarily inspired by physics and complex systems science. Contributing this direction, we recently introduced mathematical framework to describe spatiotemporal interactions neurons using lattice theory, reference paradigm for theoretical particle physics. In note, provide concise summary basics aiming be intuitive interdisciplinary community. We contextualize our methods, illustrating how readily connect parameters formulation experimental variables well-known renormalization procedures. This synopsis yields key concepts needed neural networks Such classes methods are attention-worthy an era blistering improvements numerical computations, as they can facilitate relating observation activity generative models underpinned physical principles.

Language: Английский

---

Innovating beyond electrophysiology through multimodal neural interfaces

Nature Reviews Electrical Engineering, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 16, 2024

Language: Английский

---

Development of wafer-scale multifunctional nanophotonic neural probes for brain activity mapping

Lab on a Chip, Journal Year: 2024, Volume and Issue: 24(9), P. 2397 - 2417

Published: Jan. 1, 2024

Optical techniques, such as optogenetic stimulation and functional fluorescence imaging, have been revolutionary for neuroscience by enabling neural circuit analysis with cell-type specificity. To probe deep brain regions, implantable light sources are crucial. Silicon photonics, commonly used data communications, shows great promise in creating devices complex optical systems a compact form factor compatible high volume manufacturing practices. This article reviews recent developments of wafer-scale multifunctional nanophotonic probes. The probes can be realized on 200 or 300 mm wafers commercial foundries integrate emitters photostimulation, microelectrodes electrophysiological recording, microfluidic channels chemical delivery sampling. By integrating active to the probes, denser emitter arrays, enhanced on-chip biosensing, increased ease use may realized. photonics technology makes possible highly versatile that transform experiments.

Language: Английский

---

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

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Oct. 29, 2023

High-density microelectrode arrays (MEAs) have opened new possibilities for systems neuroscience in human and non-human animals, but brain tissue motion relative to the array poses a challenge downstream analyses, particularly recordings. We introduce DREDge (Decentralized Registration of Electrophysiology Data), robust algorithm which is well suited registration noisy, nonstationary extracellular electrophysiology In addition estimating from spikes action potential (AP) frequency band, enables automated tracking at high temporal resolution local field (LFP) band. intraoperative recordings, often feature fast (period <1s) motion, correction LFP band enabled reliable recovery evoked potentials, significantly reduced single-unit spike shape variability sorting error. Applying recordings made during deep probe insertions nonhuman primates demonstrated possibility centimeters across several regions while simultaneously mapping single unit electrophysiological features. reliably delivered improved acute mouse especially those with an recent ultra-high density probe. also implemented procedure applying tens days chronic implantations mice, yielding stable despite changes neural activity experimental sessions. Together, these advances enable automated, scalable data multiple species, types, drift cases, providing foundation scientific analyses rich datasets.

Language: Английский

---

Flexible EMG arrays with integrated electronics for scalable electrode density

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

Published: July 3, 2024

Abstract Recent developments in electrode technology have demonstrated the power of flexible microelectrode arrays (FMEAs) for measuring muscle activity at high resolution. We recently introduced Myomatrix array, a FMEA optimized individual motor units (the collection fibers innervated by single neuron) [1] freely behaving animals. Although FMEAs are fundamentally changing way EMG is acquired, number recording channels limited size plug that interfaces with digital amplifier hardware and density connections on array. Increasing channel count supporting electrophysiological studies smaller animals depends two seemingly incompatible goals: reducing device while increasing channels. The solution to this increase density, which currently requiring separate headstage components be used simultaneously. In our prior devices [1], each had dedicated wire output every input, creating 1 : 1. To improve we developed novel integrating (bare-die RHD2216 chip, Intan, Inc. [6]) directly onto an FMEA. This new design reduces device’s backend footprint 74% relocates intan bare die from itself, 3.2. Our methodology combines standard microfabrication wire-bonding surface-mounted components, enabling direct integration into Serial Peripheral Interface (SPI) connection without any headstage. With initial see 3.2 but method allows using other amplifiers (Intan, Inc., USA) 12.8. findings present robust technique chip embedding custom FMEAs, applicable in-vivo electrophysiology

Language: Английский

---

RT-Sort: an action potential propagation-based algorithm for real time spike detection and sorting with millisecond latencies

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

Published: April 12, 2024

Abstract With the use of high density multi electrode recording devices, electrophysiological signals resulting from action potentials individual neurons can now be reliably detected on multiple adjacent electrodes both in vivo and vitro . Spike sorting assigns these to putative neural sources. However, until now, spike only performed after completion recording, preventing true real time usage algorithms. Utilizing unique propagation patterns along axons as fidelity sequential activations electrodes, together with a convolutional network based detection algorithm, we introduce RT-Sort (Real Time Sorting), algorithm that enables sorted within 7.5ms±1.5ms (mean±STD) waveform trough while remains ongoing. RT-Sort’s real-time capabilities enable closed loop experiments latencies comparable synaptic delay times. We show performance Multi-Electrode Arrays well Neuropixels probes exemplify functionality different types hardware configurations.

Language: Английский

---

Neuropixels probe; sex differences in brain anatomy

The Transmitter, Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

Language: Английский

---