Enhancing Assistive Technologies With Neuromorphic Computing

Advances in computational intelligence and robotics book series, Journal Year: 2024, Volume and Issue: unknown, P. 207 - 232

Published: Oct. 4, 2024

The development of intelligent neuroprosthetics, which promise to augment human brain function is vital for augmentative assistive technologies. Neuromorphic sensors and processors are particularly adept at mimicking the brain's efficient sensory processing, offering devices an advanced capability perceive interpret complex environmental stimuli. application these technologies in computer interfaces suggests a future where transformative advancements not only possible but imminent, facilitating novel methods human-computer interaction providing insights into intricate workings through AI machine learning techniques. This paper explores integration neuromorphic with brain-computer (BCIs), highlighting potential enhance revolutionize communication healthcare. However, realization computing's full within BCIs contingent upon overcoming significant technological ethical challenges.

Language: Английский

---

Spike sorting AI agent

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

Published: Feb. 12, 2025

Spike sorting is a fundamental process for decoding neural activity, involving preprocessing, spike detection, feature extraction, clustering, and validation. However, conventional methods are highly fragmented, labor-intensive, heavily reliant on expert manual curation, limiting their scalability reproducibility. This challenge has become more pressing with advances in recording technology, such as high-density Neuropixels large-scale or flexible electrodes long-term stable over months to years. The volume complexity of these datasets make curation infeasible, requiring an automated scalable solution. Here, we introduce SpikeAgent, multimodal large language model (LLM)-based AI agent that automates standardizes the entire pipeline. Unlike traditional approaches, SpikeAgent integrates multiple LLM backends, coding functions, established algorithms, autonomously performing reasoning-based decision-making real-time interaction intermediate results. It generates interpretable reports, providing transparent justifications each decision, enhancing transparency reliability. We benchmarked against human experts across various demonstrating its versatility ability achieve consistency equal to, even higher than experts. also drastically reduces expertise barrier accelerates validation time by orders magnitude. Moreover, it enables interpretability spiking data, which cannot be achieved any methods. presents paradigm shift processing signals neuroscience brain-computer interfaces, while laying ground agent-augmented science domains.

Language: Английский

---

Tracking spatial patterns and daily modulation of behavior in a natural population of the pulse-type weakly electric fish, Gymnotus omarorum

iScience, Journal Year: 2025, Volume and Issue: 28(3), P. 112018 - 112018

Published: Feb. 13, 2025

Tracking individual spatial and activity-rest patterns in natural populations is challenging because it seldom possible to monitor individual-specific traits continuously. The continuous emission of electric signals (EODs) by weakly fish provides a unique opportunity do this. We present cutting-edge technique, arrays electrodes connected low-cost amplifiers tracking algorithm, provide the identification pulse-type wild. Based only on EOD recordings individuals Gymnotus omarorum, we show that (1) there are more core than edge zones; (2) transitions into out recording sites were frequent at night, (3) resident robust nocturnal increases rate likely associated with daily variations water temperature. This experimental approach can be extended other species improve our understanding behavior, ecology, well-being environments.

Language: Английский

---

Stress monitoring using low-cost electroencephalogram devices: A systematic literature review

International Journal of Medical Informatics, Journal Year: 2025, Volume and Issue: 198, P. 105859 - 105859

Published: March 6, 2025

The use of low-cost, consumer-grade wearable health monitoring devices has become increasingly prevalent in mental research, including stress studies. While cortisol response magnitude remains the gold standard for assessment, an expanding body research employs low-cost EEG as primary tools recording biomarker data, often combined with wrist and ring-based wearables. However, technical variability among devices, particularly sensor count placement according to 10-20 Electrode Placement System, poses challenges reproducibility study outcomes. This review aims provide overview growing application machine learning techniques assessing brain function, a focus on detection. It also highlights strengths weaknesses various methods commonly used evaluates reported findings along importance. A comprehensive was conducted published studies utilizing detection their associated approaches. Searches were performed across databases Scopus, Google Scholar, ScienceDirect, Nature, PubMed, yielding 69 relevant articles analysis. selected synthesized into four thematic categories: assessment using EEG, datasets EEG-based measurement, For learning-focused studies, validation critically assessed. Study quality evaluated scored IJMEDI checklist. identified several employing monitor activity during relaxation phases, many reporting high predictive accuracy techniques. only 54% included screening prior experimentation, 58% categorized low-powered due limited sample sizes. Additionally, few validated results independent set or correlating there lack consensus data pre-processing key contributor improving model generalization accuracy. Low-cost wrist-based monitors, are utilized stress-related offering promising avenues non-invasive monitoring. significant gaps remain standardizing signal processing placement, both which critical enhancing Furthermore, sets biomarkers need more robust methodologies. Future should addressing these limitations establishing configurations improve reliability this field.

Language: Английский

---

Neuroengineering and brain-machine interfaces

Elsevier eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 325 - 357

Published: Jan. 1, 2025

Language: Английский

---

A Comprehensive Exploration of Unsupervised Classification in Spike Sorting: A Case Study on Macaque Monkey and Human Pancreatic Signals

Algorithms, Journal Year: 2024, Volume and Issue: 17(6), P. 235 - 235

Published: May 30, 2024

Spike sorting, an indispensable process in the analysis of neural biosignals, aims to segregate individual action potentials from mixed recordings. This study delves into a comprehensive investigation diverse unsupervised classification algorithms, some which, best our knowledge, have not previously been used for spike sorting. The methods encompass Principal Component Analysis (PCA), K-means, Self-Organizing Maps (SOMs), and hierarchical clustering. research draws insights both macaque monkey human pancreatic signals, providing holistic evaluation across species. Our has focused on utilization aforementioned sorting 327 detected spikes within vivo signal monkey, as well 386 vitro pancreas. was carried out by extracting statistical features these spikes. We initiated with employing unmodified normalized versions features. To enhance performance this algorithm, we also employed (PCA) reduce dimensionality data, thereby leading more distinct groupings identified K-means algorithm. Furthermore, two additional techniques, namely clustering Maps, undergone exploration demonstrated favorable outcomes types. Across all scenarios, consistent observation emerged: identification six distinctive groups spikes, each characterized shapes, sets. In regard, meticulously present thoroughly analyze experimental yielded algorithms. presentation discussion encapsulate nuances, patterns, uncovered algorithms data. By delving specifics results, aim provide nuanced understanding efficacy algorithm context

Language: Английский

---

Single-trial detection of auditory cues from the rat brain using memristors

Science Advances, Journal Year: 2024, Volume and Issue: 10(36)

Published: Sept. 4, 2024

Implantable devices hold the potential to address conditions currently lacking effective treatments, such as drug-resistant neural impairments and prosthetic control. Medical need be biologically compatible while providing enhanced performance metrics of low-power consumption, high accuracy, small size, minimal latency enable ongoing intervention in brain function. Here, we demonstrate a memristor-based processing system for single-trial detection behaviorally meaningful signals within timeframe that supports real-time closed-loop intervention. We record activity from reward center brain, ventral tegmental area, rats trained associate musical tone with reward, use memristors built-in thresholding properties detect nontrivial biomarkers local field potentials. This approach yields consistent accurate >98% maintaining power consumption low 4.14 nanowatt per channel. The efficacy our system’s capabilities process vivo data paves way chronic monitoring biomedical implants.

Language: Английский

---

Graphene Microelectrode Arrays, 4D Structured Illumination Microscopy, and a Machine Learning Spike Sorting Algorithm Permit the Analysis of Ultrastructural Neuronal Changes During Neuronal Signaling in a Model of Niemann–Pick Disease Type C

Advanced Science, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 28, 2024

Abstract Simultaneously recording network activity and ultrastructural changes of the synapse is essential for advancing understanding basis neuronal functions. However, rapid millisecond‐scale fluctuations in subtle sub‐diffraction resolution synaptic morphology pose significant challenges to this endeavor. Here, specially designed graphene microelectrode arrays (G‐MEAs) are used, which compatible with high spatial imaging across various scales as well permit temporal electrophysiological recordings address these challenges. Furthermore, alongside G‐MEAs, an easy‐to‐implement machine learning algorithm developed efficiently process large datasets collected from MEA recordings. It demonstrated that combined use (ML) spike analysis, 4D structured illumination microscopy (SIM) enables monitoring impact disease progression on hippocampal neurons treated intracellular cholesterol transport inhibitor mimicking Niemann–Pick type C (NPC), show boutons, compared untreated controls, significantly increase size, leading a loss signaling capacity.

Language: Английский

---

A comprehensive review of spike sorting algorithms in neuroscience

Progress in Medical Devices, Journal Year: 2024, Volume and Issue: unknown

Published: June 30, 2024

Spike sorting plays a pivotal role in neuroscience, serving as crucial step of separating electrical signals recorded from multiple neurons to further analyze neuronal interactions. This process involves that originate neurons, through devices like electrode arrays. is very important link the field brain-computer interfaces. The objective spike algorithm (SSA) distinguish behavior one or more background noise using waveforms captured by brain-embedded electrodes. article starts steps conventional SSA and divides into three steps: detection, feature extraction, clustering. It outlines prevalent algorithms for each phase before delving two emerging technologies: template matching deep learning-based methods. discussion on learning subdivided approaches: end-to-end solution, steps, spiking neural networks-based solutions. Finally, it elaborates future challenges development trends SSAs.

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

---