A5084081295- Neural dynamics and brain function
- Neuroscience and Neural Engineering
- Advanced Memory and Neural Computing
- Neuroscience and Neuropharmacology Research
- EEG and Brain-Computer Interfaces
- Neural Networks and Applications
- Photoreceptor and optogenetics research
- Functional Brain Connectivity Studies
- Visual perception and processing mechanisms
- stochastic dynamics and bifurcation
- Neurobiology and Insect Physiology Research
- Zebrafish Biomedical Research Applications
- Advanced Chemical Sensor Technologies
- Complex Systems and Time Series Analysis
- Visual Attention and Saliency Detection
- Blind Source Separation Techniques
- Face Recognition and Perception
- Olfactory and Sensory Function Studies
- Photoacoustic and Ultrasonic Imaging
- Image and Signal Denoising Methods
- Fractal and DNA sequence analysis
- Nonlinear Dynamics and Pattern Formation
- Advanced Vision and Imaging
- Ion channel regulation and function
- Ferroelectric and Negative Capacitance Devices
Babeș-Bolyai University
2023-2025
Technical University of Cluj-Napoca
2004-2024
Transylvania University
2021-2024
Romanian Institute of Science and Technology
2011-2023
Max Planck Institute for Brain Research
2008-2012
Max Planck Society
2009
Frankfurt Institute for Advanced Studies
2005-2007
Goethe University Frankfurt
2006
Due to the Heisenberg-Gabor uncertainty principle, finite oscillation transients are difficult localize simultaneously in both time and frequency. Classical estimators, like short-time Fourier transform or continuous-wavelet optimize either temporal frequency resolution, find a suboptimal tradeoff. Here, we introduce spectral estimator enabling time-frequency super-resolution, called superlet, that uses sets of wavelets with increasingly constrained bandwidth. These combined geometrically...
While animals readily adjust their behavior to adapt relevant changes in the environment, neural pathways enabling these remain largely unknown. Here, using multiphoton imaging, we investigate whether feedback from piriform cortex olfactory bulb supports such behavioral flexibility. To this end, engage head-fixed male mice a multimodal rule-reversal task guided by and auditory cues. Both odor and, surprisingly, sound cues trigger responses cortical bulbar axons which precede report....
Abstract When computing a cross‐correlation histogram, slower signal components can hinder the detection of faster components, which are often in research focus. For example, precise neuronal synchronization co‐occurs with slow co‐variation rate responses. Here we present method – dubbed scaled correlation analysis that enables isolation histogram fast components. The computes correlations only on small temporal scales (i.e. short segments signals such as 25 ms), resulting removal than those...
Neuronal mechanisms underlying beta/gamma oscillations (20–80 Hz) are not completely understood. Here, we show that in vivo the cat visual cortex sometimes exhibit remarkably stable frequency even when inputs fluctuate dramatically. Enhanced stability is associated with stronger measured individual units and larger power local field potential. Simulations of neuronal circuitry demonstrate membrane properties inhibitory interneurons strongly determine characteristics emergent oscillations....
We present a method that estimates the strength of neuronal oscillations at cellular level, relying on autocorrelation histograms computed spike trains. The delivers number, termed oscillation score, degree to which neuron is oscillating in given frequency band. Moreover, it can also reliably identify and band, independently other bands, thus handle superimposed multiple scales (theta, alpha, beta, gamma, etc.). relatively simple fast. It cope with low number spikes, converging exponentially...
Abstract Objective DEPDC 5 was identified as a major genetic cause of focal epilepsy with deleterious mutations found in wide range inherited forms epilepsy, associated malformation cortical development certain cases. Identification frameshift, truncation, and deletion implicates haploinsufficiency the etiology epilepsy. is component GATOR 1 complex, acting negative regulator mTOR signaling. Methods Zebrafish represents vertebrate model suitable for analysis drug screening epilepsy‐related...
Spike sorting is the process of grouping spikes distinct neurons into their respective clusters. Most frequently, this performed by relying on similarity features extracted from spike shapes. In spite recent developments, current methods have yet to achieve satisfactory performance and many investigators favour manually, even though it an intensive undertaking that requires prolonged allotments time. To automate process, a diverse array machine learning techniques has been applied. The these...
We investigated spontaneous activity and excitability in large networks of artificial spiking neurons. compared three different neuron models: integrate-and-fire (IF), regular-spiking (RS), resonator (RES). First, we show that models have frequency-dependent response properties, yielding differences excitability. Then, investigate the responsiveness these to a single afferent inhibitory/excitatory spike calibrate total synaptic drive such they would exhibit similar peaks postsynaptic...
Abstract Focal cortical dysplasia, hemimegalencephaly and tubers are paediatric epileptogenic malformations of development (MCDs) frequently pharmacoresistant mostly treated surgically by the resection epileptic cortex. Availability samples has allowed significant mechanistic discoveries directly from human material. Causal brain somatic or germline mutations in AKT/PI3K/DEPDC5/MTOR genes have been identified. GABAA-mediated paradoxical depolarization, related to altered chloride (Cl−)...
Abstract Baseline normalization procedures are essential for the analysis of brain activity. These use statistics a reference (baseline) period to normalize data along entire trial (baseline and stimulus periods). A very popular procedure is pseudo z ‐scoring, traditionally applied time–frequency spectral power estimates, where it was recently shown generate positive bias. Bias thought arise because outliers stemming from skewed distribution values. Here we challenge this view causally show...
A key strategy to enable training of deep neural networks is use non-saturating activation functions reduce the vanishing gradient problem. Popular choices that saturate only in negative domain are rectified linear unit (ReLU), its smooth, non-linear variant, Softplus, and exponential units (ELU SELU). Other across entire real domain, like parametric ReLU (PReLU). Here we introduce a nonlinear function called Soft++ extends PReLU parametrizing slope exponent. We test identical network...
Eye-tracking is a method of recording the location gaze as well pupil diameter (dilation) during active visual behavior. Due to blinking and noise in system, these signals are often briefly "lost", leading missing data. Here, we aim analyze accuracy six interpolation methods complete values from Data type estimation that constructs new data existing, neighboring values. Having possibility choose different types methods, question which most suitable for applied linear method, previous...
Functional magnetic resonance imaging (fMRI) provides insights into cognitive processes with significant clinical potential. However, delays in brain region communication and dynamic variations are often overlooked functional network studies. We demonstrate that networks extracted from fMRI cross-correlation matrices, considering time lags between signals, show remarkable reliability when focusing on statistical distributions of properties. This reveals a robust connectivity pattern,...
Mechanisms of explicit object recognition are often difficult to investigate and require stimuli with controlled features whose expression can be manipulated in a precise quantitative fashion. Here, we developed novel method (called "Dots"), for generating visual stimuli, which is based on the progressive deformation regular lattice dots, driven by local contour information from images objects. By applying progressively larger lattice, latter conveys more about target object. Stimuli...
The electroencephalography (EEG) data records vast amounts of human cerebral activity yet is still reviewed primarily by readers. Most the times, contaminated with non-cerebral originated signals, called artifacts, which could be very difficult to visually detect and, undiscovered, damage neural information analysis. purpose our work artifacts identifying most relevant features, both in temporal and frequency domains, train various supervised learning algorithms: Decision Trees, SVM KNN,...
Brain oscillations are thought to subserve important functions by organizing the dynamical landscape of neural circuits. The expression such in signals is usually evaluated using time-frequency representations (TFR), which resolve oscillatory processes both time and frequency. While a vast number methods exist compute TFRs, there often no objective criterion decide one better. In feature-rich data, as that recorded from brain, sources noise unrelated abound contaminate results. impact these...
The main objective of this paper is the time-frequency analysis EEG signal captured in a cognitive task (i.e. object recognition) performed by human subjects. We investigate whether power spectral density gamma frequency range can be used to classify outcome recognition seen, unseen, uncertain). signals were acquired and analyzed from 128 electrodes located on all parts brain. Power features are extracted for classification support vector machine (SVM), K-Nearest Neighbor (KNN) Artificial...
Abstract Time-frequency analysis is ubiquitous in many fields of science. Due to the Heisenberg-Gabor uncertainty principle, a single measurement cannot estimate precisely location finite oscillation both time and frequency. Classical spectral estimators, like short-time Fourier transform (STFT) or continuous-wavelet (CWT) optimize either temporal frequency resolution, find tradeoff that suboptimal dimensions. Following concepts from optical super-resolution, we introduce new estimator...
Abstract Recognising objects is a vital skill on which humans heavily rely to respond quickly and adaptively their environment. Yet, we lack full understanding of the role visual information sampling plays in this process, its relation individual’s priors. To bridge gap, eye-movements 18 adult participants were recorded during free-viewing object-recognition task using Dots stimuli 1 . Participants viewed one three orders: from most visible least ( Descending ), Ascending or randomised order...
The Continuous Wavelet Transform (CWT) provides a multi-resolution representation of signal by scaling mother wavelet and convolving it with the signal. scalogram (squared modulus CWT) then represents spread signal's energy as function time scale. has constant relative temporal resolution but, scale is compressed (frequency increased), loses frequency resolution. To compensate for this, recently-introduced superlets geometrically combine set wavelets increasing to achieve time-frequency...