Search

Abstract The electroencephalogram (EEG) is a tool for diagnosing epilepsy; by analyzing it, neurologists can identify alterations in brain activity associated with epilepsy. However, this task is not always easy to perform because of the duration of the EEG or the subjectivity of the specialist in detecting alterations. Aim To propose the use of an epileptic spike detector based on a matched filter and a neural network for supporting the diagnosis of epilepsy through a tool capable of automatically detecting spikes in pediatric EEGs. Results Automatic detection of spikes from an EEG waveform involved the creation of an epileptic spike template. The template was used in order to detect spikes by using a matched filter, and each spike detected was confirmed by a Neural Network to improve sensitivity and specificity. Thus, the detector developed achieved a sensitivity of 99.96% which is better than the range of what has been reported in the literature (82.68% and 94.4%), and a specificity of 99.26%, improving the specificity found in the best-reviewed studies. Conclusions Considering the results obtained in the evaluation, the solution becomes a promising alternative to support the automatic identification of epileptic spikes by neurologists.

Spike is one of the crop yield organs in wheat plants. Determination of the phenological stages, including heading time point (HTP), and area of spike from non-invasive phenotyping images provides the necessary information for the inference of growth-related traits. The algorithm previously developed by Qiongyan et al. for spike detection in 2-D images turns out to be less accurate when applied to the European cultivars that produce many more leaves. Therefore, we here present an improved and extended method where (i) wavelet amplitude is used as an input to the Laws texture energy-based neural network instead of original grayscale images and (ii) non-spike structures (e.g., leaves) are subsequently suppressed by combining the result of the neural network prediction with a Frangi-filtered image. Using this two-step approach, a 98.6% overall accuracy of neural network segmentation based on direct comparison with ground-truth data could be achieved. Moreover, the comparative error rate in spike HTP detection and growth correlation among the ground truth, the algorithm developed by Qiongyan et al., and the proposed algorithm are discussed in this paper. The proposed algorithm was also capable of significantly reducing the error rate of the HTP detection by 75% and improving the accuracy of spike area estimation by 50% in comparison with the Qionagyan et al. method. With these algorithmic improvements, HTP detection on a diverse set of 369 plants was performed in a high-throughput manner. This analysis demonstrated that the HTP of 104 plants (comprises of 57 genotypes) with lower biomass and tillering range (e.g., earlier-heading types) were correctly determined. However, fine-tuning or extension of the developed method is required for high biomass plants where spike emerges within green bushes. In conclusion, our proposed method allows significantly more reliable results for HTP detection and spike growth analysis to be achieved in application to European cultivars with earlier-heading types.

This paper extends the class of continuous-time signals that can be perfectly reconstructed by developing a theory for the sampling and exact reconstruction of streams of short pulses with unknown shapes. The single pulse is modelled as the delayed version of a wavelet-sparse signal, which is normally not band limited. As the delay can be an arbitrary real number, it is hard to develop an exact sampling result for this type of signals. We achieve the exact reconstruction of the pulses by using only the knowledge of the Fourier transform of the signal at specific frequencies. We further introduce a multi-channel acquisition system which uses a new family of compact-support sampling kernels for extracting the Fourier information from the samples. The shape of the kernel is independent of the wavelet basis in which the pulse is sparse and hence the same acquisition system can be used with pulses which are sparse on different wavelet bases. By exploiting the fact that pulses have short duration and that the sampling kernels have compact support, we finally propose a local and sequential algorithm to reconstruct streaming pulses from the samples.

International audience; Epileptiform discharges in interictal electroencephalography (EEG) form the mainstay of epilepsy diagnosis and localization of seizure onset. Visual analysis is rater-dependent and time consuming, especially for long-term recordings, while computerized methods can provide efficiency in reviewing long EEG recordings. This paper presents a machine learning approach for automated detection of epileptiform discharges (spikes). The proposed method first detects spike patterns by calculating similarity to a coarse shape model of a spike waveform and then refines the results by identifying subtle differences between actual spikes and false detections. Pattern classification is performed using Support Vector Machines (SVM) in a low dimensional space on which the original waveforms are embedded by Locality Preserving Projections (LPP). The automatic detection results are compared to experts' manual annotations (101 spikes) on a whole-night sleep EEG recording. The high sensitivity (97%) and the low false positive rate (0.1 min-1), calculated by intra-patient cross-validation, highlight the potential of the method for automated interictal EEG assessment.

Real-time neural spike detection is an important step in understanding neurological activities and developing brain-silicon interfaces. Recent approaches exploit minimally invasive sensing techniques based on implanted complementary metal-oxide semiconductors (CMOS) multi transistors arrays (MTAs) that limit the damage of the neural tissue and provide high spatial resolution. Unfortunately, MTAs result in low signal-to-noise ratios due to the weak capacitive coupling between the nearby neurons and the sensor and the high noise power coming from the analog front-end. In this paper we investigate the performance achievable by using spike detection algorithms for MTAs, based on some variants of the smoothed non-linear energy operator (SNEO). We show that detection performance benefits from the correlation of the signals detected by the MTA pixels, but degrades when a high firing rate of neurons occurs. We present and compare different approaches and noise estimation techniques for the SNEO, aimed at increasing the detection accuracy at low SNR and making it less dependent on neurons firing rates. The algorithms are tested by using synthetic neural signals obtained with a modified version of NEUROCUBE generator. The proposed approaches outperform the SNEO, showing a more than 20% increase on averaged sensitivity at 0 dB and reduced dependence on the neuronal firing rate.

Chronic extracellular recording is the use of implanted electrodes to measure the electrical activity of nearby neurons over a long period of time. It presents an unparalleled view of neural activity over a broad range of time scales, offering sub-millisecond resolution of single action potentials while also allowing for continuous recording over the course of many months. These recordings pick up a rich collection of neural phenomena -- spikes, ripples, and theta oscillations, to name a few -- that can elucidate the activity of individual neurons and local circuits. However, this also presents an interesting challenge for data analysis. Chronic extracellular recordings contain overlapping signals from multiple sources, requiring these signals to be detected and classified before they can be properly analyzed. The combination of fine temporal resolution with long recording durations produces large datasets, requiring efficient algorithms that can operate at scale. In this thesis, I consider the problem of spike sorting: detecting spikes (the extracellular signatures of individual neurons' action potentials) and clustering them according to their putative source. First, I introduce a sparse deconvolution approach to spike detection, which seeks to detect spikes and represent them as the linear combination of basis waveforms. This approach is able to separate overlapping spikes without the need for source templates, and produces an output that can be used with a variety of clustering algorithms. Second, I introduce a clustering algorithm based around a mixture of drifting t-distributions. This model captures two features of chronic extracellular recordings -- cluster drift over time and heavy-tailed residuals in the distribution of spikes -- that are missing from previous models. This enables us to reliably track individual neurons over longer periods of time. I will also show that this model produces more accurate estimates of classification error, which is an important component to proper interpretation of the spike sorting output. Finally, I present a few theoretical results that may assist in the efficient implementation of sparse deconvolution.

Uncertainty relations play a prominent role in signal processing, stating that a signal can not be simultaneously concentrated in the two related domains of the corresponding phase space. In particular, a new uncertainty principle for the affine group, which is directly related to the wavelet transform has lead to a new minimizing waveform. In this thesis, a frame construction is proposed which leads to approximately tight frames based on this minimizing waveform. Frame properties such as the diagonality of the frame operator as well as lower and upper frame bounds are analyzed. Additionally, three applications of such frame constructions are introduced: inpainting of missing audio data, detection of neuronal spikes in extracellular recorded data and peak detection in MALDI imaging data.

ObjectiveThe process of manually marking up epileptic spikes for simultaneous electroencephalogram (EEG) and resting state functional MRI (rsfMRI) analysis in epilepsy studies is a tedious and subjective task for a human expert. The aim of this study was to evaluate whether automatic EEG spike detection can facilitate EEG-rsfMRI analysis, and to assess its potential as a clinical tool in epilepsy.MethodsWe implemented a fast algorithm for detection of uniform interictal epileptiform discharges (IEDs) in one-hour scalp EEG recordings of 19 refractory focal epilepsy datasets (from 16 patients) who underwent a simultaneous EEG-rsfMRI recording. Our method was based on matched filtering of an IED template (derived from human markup) used to automatically detect other ‘similar’ EEG events. We comprehensively compared simultaneous EEG-rsfMRI results between automatic IED detection and standard analysis with human EEG markup only.ResultsIn contrast to human markup, automatic IED detection takes a much shorter time to detect IEDs and export an output text file containing spike timings. In 13/19 focal epilepsy cases, statistical EEG-rsfMRI maps based on automatic spike detection method were comparable with human markup, and in 6/19 focal epilepsy cases it revealed additional brain regions not seen with human EEG markup. Additional events detected by our automated method independently revealed similar patterns of activation to a human markup. Overall, automatic IED detection provides greater statistical power in EEG-rsfMRI analysis compared to human markup in a short timeframe.ConclusionsAutomatic spike detection is a simple and fast method that can reproduce comparable and, in some cases, even superior results compared to the common practice of manual EEG markup in EEG-rsfMRI analysis of epilepsy.SignificanceOur study shows that IED detection algorithms can be effectively used in epilepsy clinical settings. This work further helps in translating EEG-rsfMRI research into a fast, reliable and easy-to-use clinical tool for epileptologists. Our IED detection approach will be publicly available as a MATLAB package at: https://github.com/omidvarnia/Automatic_focal_spike_detection.HighlightsAutomatic spike detection increases the number of detected uniform epileptic interictal discharges and enhances statistical power of EEG-rsfMRI inter-subject variability maps,Automatic spike detection can identify additional activated brain regions with presumed epileptogenic focus not seen in standard analysis based on human markup,Automatic spike detection can shorten the IED identification process.

Objective Longitudinal observation of single unit neural activity from large numbers of cortical neurons in awake and mobile animals is often a vital step in studying neural network behaviour and towards the prospect of building effective brain-machine interfaces (BMIs). These recordings generate enormous amounts of data for transmission and storage, and typically require offline processing to tease out the behaviour of individual neurons. Our aim was to create a compact system capable of: (1) reducing the data bandwidth by circa 2 to 3 orders of magnitude (greatly improving battery lifetime and enabling low power wireless transmission in future versions); (2) producing real-time, low-latency, spike sorted data; and (3) long term untethered operation. Approach We have developed a headstage that operates in two phases. In the short training phase a computer is attached and classic spike sorting is performed to generate templates. In the second phase the system is untethered and performs template matching to create an event driven spike output that is logged to a micro-SD card. To enable validation the system is capable of logging the high bandwidth raw neural signal data as well as the spike sorted data. Main results The system can successfully record 32 channels of raw neural signal data and/or spike sorted events for well over 24 h at a time and is robust to power dropouts during battery changes as well as SD card replacement. A 24 h initial recording in a non-human primate M1 showed consistent spike shapes with the expected changes in neural activity during awake behaviour and sleep cycles. Significance The presented platform allows neural activity to be unobtrusively monitored and processed in real-time in freely behaving untethered animals-revealing insights that are not attainable through scheduled recording sessions. This system achieves the lowest power per channel to date and provides a robust, low-latency, low-bandwidth and verifiable output suitable for BMIs, closed loop neuromodulation, wireless transmission and long term data logging.

Modern electroceuticals are bound to employ the usage of electrical high frequency (130–180 Hz) stimulation carried out under closed loop control, most prominent in the case of movement disorders. However, particular challenges are faced when electrical recordings of neuronal tissue are carried out during high frequency electrical stimulation, both in-vivo and in-vitro. This stimulation produces undesired artifacts and can render the recorded signal only partially useful. The extent of these artifacts is often reduced by temporarily grounding the recording input during stimulation pulses. In the following study, we quantify the effects of this method, “blanking,” on the spike count and spike train statistics. Starting from a theoretical standpoint, we calculate a loss in the absolute number of action potentials, depending on: width of the blanking window, frequency of stimulation, and intrinsic neuronal activity. These calculations were then corroborated by actual high signal to noise ratio (SNR) single cell recordings. We state that, for clinically relevant frequencies of 130 Hz (used for movement disorders) and realistic blanking windows of 2 ms, up to 27% of actual existing spikes are lost. We strongly advice cautioned use of the blanking method when spike rate quantification is attempted. Impact statement Blanking (artifact removal by temporarily grounding input), depending on recording parameters, can lead to significant spike loss. Very careful use of blanking circuits is advised.

International audience; Epilepsy is one of the diseases that are more subject to consultation in neurological clinics. To help neurologists to accurately diagnose this disease, several technological tools have been developed. Electroencephalography (EEG) of scalp or deep is a signal acquisition tool from electrical discharges of the brain areas. These signals are often accompanied by transient events commonly called interictal paroxystic events (IPE) or spikes of short durations. Analysis of these IPE could help with the diagnosis of drug-resistant epilepsy. With this intention, we will first of all seek to detect IPE, by separating them from the basic activity of signal EEG. In this paper, we propose spike detection method based on Smoothed Nonlinear Energy Operator (SNEO) using adaptive threshold. Then we will implement a new approach using principal components analysis (PCA) before classification to separate the events detected according to their morphologies. The objective in the long term is to characterize their space-time distribution over all the duration of the EEG signal. ©ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering 2018.

This paper reports an integrated 64-channel neural spike recording sensor, together with all the circuitry to process and configure the channels, process the neural data, transmit via a wireless link the information and receive the required instructions. Neural signals are acquired, filtered, digitized and compressed in the channels. Additionally, each channel implements an auto-calibration algorithm which individually configures the transfer characteristics of the recording site. The system has two transmission modes; in one case the information captured by the channels is sent as uncompressed raw data; in the other, feature vectors extracted from the detected neural spikes are released. Data streams coming from the channels are serialized by the embedded digital processor. Experimental results, including in vivo measurements, show that the power consumption of the complete system is lower than 330 μW.

International audience; The acquisition of EEG signals can be done during days. In this paper, the signals of depth Stereo-Electroencephalography (SEEG) are used. In SEEG signal, several inter-ictal paroxystic events (IPE) are found, they appear between the crises. To analyze their distribution, they should initially be detected by separating them from the basic activity. After the detection step, we implement a clustering process to distinguish IPE or spike according to their morphology. Results of detection and clustering would characterize a link between space-time distribution of IPE and the arrival of the crises. It would be thus a contribution on the diagnosis of the partial epilepsies. © 2016 IEEE.

The subthalamic nucleus (STN) plays a central role in movement actuation and manifestation of movement disorders (i.e., tremor, rigidity, akynesia and postural instability) in Parkinson's disease (PD) patients. Moreover, it has been recently revealed that an opportune electrical stimulation of the STN, called deep brain stimulation (DBS), can strongly contribute to the annihilation of the PD-related motor disorders. Currently, a great effort is made both in Medicine, Neurosciences and Engineering for understanding and modeling in details how the STN works, how PD determines its pathological behavior and DBS restores the correct motor function. The paper is organized in two parts. Firstly some stochastic properties of the STN electrical activity are obtained by analyzing a preliminary set of experimental data coming from microelectrode recordings (MERs) in two PD patients who underwent the surgical implantation of DBS electrodes. Then, a nonlinear, stochastic, continuous-state model describing the global electrical behavior of the STN in PD patients is proposed. It is inspired by the fundamental physiologic features of the subthalamic cells and a fictitious vector state is introduced to represent the main dynamics. Its numerical parameters and stochastic properties are chosen by fitting the available data.

We investigate methods for detection of rapid workload increases (load spikes) for cloud workloads. Such rapid and unexpected workload spikes are a main cause for poor performance or even crashing applications as the allocated cloud resources become insufficient. To detect the spikes early is fundamental to perform corrective management actions, like allocating additional resources, before the spikes become large enough to cause problems. For this, we propose a number of methods for early spike detection, based on established techniques from adaptive signal processing. A comparative evaluation shows, for example, to what extent the different methods manage to detect the spikes, how early the detection is made, and how frequently they falsely report spikes.

Information in the nervous system is coded as firing patterns of electrical signals called action potential or spike so an essential step in analysis of neural mechanism is detection of action potentials embedded in the neural data. There are several methods proposed in the literature for such a purpose. In this paper a novel method based on empirical mode decomposition (EMD) has been developed. EMD is a decomposition method that extracts oscillations with different frequency range in a waveform. The method is adaptive and no a-priori knowledge about data or parameter adjusting is needed in it. The results for simulated data indicate that proposed method is comparable with wavelet based methods for spike detection. For neural signals with signal-to-noise ratio near 3 proposed methods is capable to detect more than 95% of action potentials accurately., {"references":["Kyung H. Kim, Sung J. Kim, \"A Wavelet-Based Method for Action Potential Detection From Extracellular Neural Signal Recording With Low Signal-to-Noise Ratio,\" IEEE Trans. biomed. eng., vol. 50, no. 8, pp. 999-1011, 2003.","I. Obeid and P. Wolf, \"Evaluation of spike detection algorithms for a brain-machine interface application,\" IEEE Trans Biomed Eng, vol. 51, no. 6, pp. 905–911, 2004.","R. Chandra and L.M. Optican, \"Detection, classification, and superposition resolution of action potentials in multiunit single-channel recordings by an online real-time neural network,\" IEEE Trans. Biomed. Eng., vol. 44, pp. 403–412, 1997.","Sunghan Kim, James McNames, \"Automatic spike detection based on adaptive template matching for extracellular neural recordings,\" J. Neurosci. Methods, vol.165, pp. 165–174, 2007.","Robert J. Brychta et al., \"Wavelet Methods for Spike Detection in Mouse Renal Sympathetic Nerve Activity,\" IEEE Trans. Biomed Eng., vol. 54(1), pp. 82–93, January 2007.","Shahjahan Shahid, Jacqueline Walker, and Leslie S Smith, \"A new spike detection algorithm for extracellular neural recordings,\" IEEE Trans. Biomed Eng., vol. 57, pp.853-866, 2010.","Lewicki. M. S, \"A Review of Methods for Spike Sorting: The Detection and Classification of Neural Action Potentials,\" Network, Vol. 9, pp. 53- 78, 1998.","Huang, N., Z. Shen, S. Long, M. Wu, E. Shih, Q. Zheng, C. Tung, and H. Liu, \"The empirical mode decomposition method and the Hilbert spectrum for non-stationary time series analysis,\" Proceedings of the Royal Society of London, vol. 454, pp. 903–995, 1998.","P. Flandrin, G. Rilling, and P. Gonc¸alves, \"Empirical mode decomposition as a filter bank,\" IEEE Signal Processing Letters, vol. 11, no. 2, part 1, pp. 112–114, 2004.\n[10]\tZhaohua Wu, Norden E. Huang, \"A study of the characteristics of white noise using the empirical mode decomposition method,\" Royal Society of London Proceedings Series A, vol. 460, Issue 2046, pp. 1597-1611, 2004.\n[11]\tKais Khaldi et al., \"Speech Enhancement via EMD,\" Eurasip Journal on Advances in Signal Processing,vol. 8, pp. 1-8, 2008.\n[12]\t D. L. Donoho and I. M. Johnstone, \"Adapting to unknown smoothness via wavelet shrinkage,\" Journal of the American Statistical Association, vol. 90, no. 432, pp. 1200–1424, 1995.\n[13]\tLeslie S. Smith, Nhamoinesu Mtetwa, \"A tool for synthesizing spike trains with realistic interference,\" J. Neurosci. Methods, vol. 159, pp. 170–180, 2007."]}

Simultaneous recordings of single and multi-unit neural signals from multiple cortical areas in the brain are a vital tool for gaining more understanding of the operating mechanism of the brain as well as for developing Brain Machine Interfaces. Monitoring the activity levels of hundreds or even thousands of neurons can lead to reliable decoding of brain signals for controlling prosthesis of multiple degrees of freedom and different functionalities. With the advancement of high density microelectrode arrays, the craving of neuroscience research to record the activity of thousands of neurons is achievable. Recently CMOS-based Micro-electrode Arrays MEAs featuring high spatial and temporal resolution have been reported. The augmentation in the number of recording sites carries different challenges to the neural signal processing system. The primary challenge is the massive increase in the incoming data that needs to be transmitted and processed in real time. Data reduction based on the sparse nature of the neural signals with respect to time becomes essential. The dissertation presents the design of a neural spike-based data reduction platform that can handle a few thousands of channels on Field Programmable Gate Arrays (FPGAs), making use of their massive parallel processing capabilities and reconfigurability. For Standalone implementation the spike detector core uses Finite State Machines (FSMs) to control the interface with the data acquisition as well as sending the spike waveforms to a common output FIFO. The designed neural signal processing platform integrates the application of high-speed serial Multi-Gigabit transceivers on FPGAs to allow massive data transmission in real time. It also provides a design for autonomous threshold setting for each channel.

Estuaries and coastal areas contain increasingly exploited resources that need to be monitored, managed and protected efficiently and effectively. This requires access to reliable and timely data and management decisions must be based on analysis of collected data to avoid or limit negative impacts. Visually supported multi-modal sensing and data fusion offer attractive possibilities for such arduous tasks. In this paper, we demonstrate how an in-situ sensor network can be enhanced with the use of contextual image data. We assimilate and alter a state-of-the-art background modelling technique from the image processing domain in order to detect turbidity spikes in water quality sensor measurements automatically. We then combine this with visual sensing to identify abnormal events that are not caused by local activities. The system can potentially assist those charged with monitoring large scale ecosystems, combining real-time analytics with improved efficiency and effectiveness.

Automatic interpretation of the EEG has so far been faced with significant difficulties because of a large amount of spatial as well as temporal information contained in the EEG, continuous fluctuation of the background activity depending on changes in the subject's vigilance and attention level, the occurrence of paroxysmal activities such as spikes and spike-and-slow-waves, contamination of the EEG with a variety of artefacts and the use of different recording electrodes and montages. Therefore, previous attempts of automatic EEG interpretation have been focussed only on a specific EEG feature such as paroxysmal abnormalities, delta waves, sleep stages and artefact detection. As a result of a long-standing cooperation between clinical neurophysiologists and system engineers, we report for the first time on a comprehensive, computer-assisted, automatic interpretation of the adult waking EEG. This system analyses the background activity, intermittent abnormalities, artefacts and the level of vigilance and attention of the subject, and automatically presents its report in written form. Besides, it also detects paroxysmal abnormalities and evaluates the effects of intermittent photic stimulation and hyperventilation on the EEG. This system of automatic EEG interpretation was formed by adopting the strategy that the qualified EEGers employ for the systematic visual inspection. This system can be used as a supplementary tool for the EEGer's visual inspection, and for educating EEG trainees and EEG technicians.

This paper reports a programmable 400 μm pitch neural spike recording channel, fabricated in a 130 nm standard CMOS technology, which implements amplification, filtering, digitization, analog spike detection plus feature extraction, and self-calibration functionalities. It can operate in two different output modes: 1) signal tracking, in which the neural signal is sampled and transmitted as raw data; and 2) feature extraction, in which the spikes of the neural signal are detected and encoded by piece-wise linear curves. Additionally, the channel offers a foreground calibration procedure in which the amplification gain and the passband of the embedded filter can be self-adjusted. The amplification stage obtains a noise efficiency factor of 2.16 and an input referred noise of 2.84 μVrms over a nominal bandwidth of 167 Hz-6.9 kHz. The channel includes a reconfigurable 8-bit analog-to-digital converter combined with a 3-bit controlled programmable gain amplifier for adjusting the input signal to the full scale range of the converter. This combined block achieves an overall energy consumption per conversion of 102 fJ at 90 kS/s. The energy consumed by the circuit elements which are strictly related to the digitization process is 14.12 fJ at the same conversion rate. The complete channel consumes 2.8 μW at 1.2 V voltage supply when operated in the signal tracking mode, and 3.1 μW when the feature extraction mode is enabled.

The electroencephalogram (EEG) is a medical technology that is used in the monitoring of the brain and in the diagnosis of many neurological illnesses. Although coarse in its precision, the EEG is a non-invasive tool that requires minimal set-up times, and is suitably unobtrusive and mobile to allow continuous monitoring of the patient, either in clinical or domestic environments. Consequently, the EEG is the current tool-of-choice with which to continuously monitor the brain where temporal resolution, ease-of- use and mobility are important. Traditionally, EEG data are examined by a trained clinician who identifies neurological events of interest. However, recent advances in signal processing and machine learning techniques have allowed the automated detection of neurological events for many medical applications. In doing so, the burden of work on the clinician has been significantly reduced, improving the response time to illness, and allowing the relevant medical treatment to be administered within minutes rather than hours. However, as typical EEG signals are of the order of microvolts (μV ), contamination by signals arising from sources other than the brain is frequent. These extra-cerebral sources, known as artefacts, can significantly distort the EEG signal, making its interpretation difficult, and can dramatically disimprove automatic neurological event detection classification performance. This thesis therefore, contributes to the further improvement of auto- mated neurological event detection systems, by identifying some of the major obstacles in deploying these EEG systems in ambulatory and clinical environments so that the EEG technologies can emerge from the laboratory towards real-world settings, where they can have a real-impact on the lives of patients. In this context, the thesis tackles three major problems in EEG monitoring, namely: (i) the problem of head-movement artefacts in ambulatory EEG, (ii) the high numbers of false detections in state-of-the-art, automated, epileptiform activity detection systems and (iii) false detections in state-of-the-art, automated neonatal seizure detection systems. To accomplish this, the thesis employs a wide range of statistical, signal processing and machine learning techniques drawn from mathematics, engineering and computer science. The first body of work outlined in this thesis proposes a system to automatically detect head-movement artefacts in ambulatory EEG and utilises supervised machine learning classifiers to do so. The resulting head-movement artefact detection system is the first of its kind and offers accurate detection of head-movement artefacts in ambulatory EEG. Subsequently, addtional physiological signals, in the form of gyroscopes, are used to detect head-movements and in doing so, bring additional information to the head- movement artefact detection task. A framework for combining EEG and gyroscope signals is then developed, offering improved head-movement arte- fact detection. The artefact detection methods developed for ambulatory EEG are subsequently adapted for use in an automated epileptiform activity detection system. Information from support vector machines classifiers used to detect epileptiform activity is fused with information from artefact-specific detection classifiers in order to significantly reduce the number of false detections in the epileptiform activity detection system. By this means, epileptiform activity detection which compares favourably with other state-of-the-art systems is achieved. Finally, the problem of false detections in automated neonatal seizure detection is approached in an alternative manner; blind source separation techniques, complimented with information from additional physiological signals are used to remove respiration artefact from the EEG. In utilising these methods, some encouraging advances have been made in detecting and removing respiration artefacts from the neonatal EEG, and in doing so, the performance of the underlying diagnostic technology is improved, bringing its deployment in the real-world, clinical domain one step closer.

Invasive brain-computer interfaces (BCIs) may prove to be a useful rehabilitation tool for severely disabled patients. Although some systems have shown to work well in restricted laboratory settings, their usefulness must be tested in less controlled environments. Our objective was to investigate if a specific motor task could reliably be detected from multi-unit intra-cortical signals from freely moving animals. Four rats were trained to hit a retractable paddle (defined as a "hit"). Intra-cortical signals were obtained from electrodes placed in the primary motor cortex. First, the signal-to-noise ratio was increased by wavelet denoising. Action potentials were then detected using an adaptive threshold, counted in three consecutive time intervals and were used as features to classify either a "hit" or a "no-hit" (defined as an interval between two "hits"). We found that a "hit" could be detected with an accuracy of 75 ± 6% when wavelet denoising was applied whereas the accuracy dropped to 62 ± 5% without prior denoising. We compared our approach with the common daily practice in BCI that consists of using a fixed, manually selected threshold for spike detection without denoising. The results showed the feasibility of detecting a motor task in a less restricted environment than commonly applied within invasive BCI research. peerReviewed

In this thesis we bring together advanced mathematical tools and signal processing methods to address three very important problems in neuroscience: neural action potential (spike) detection, neural spike sorting and directly neural coding. Starting from extracellular neural recordings, we first address the question of spike detection. The spike time occurrences appear (as irregularities) explicitly in the distributional derivatives of the neural signal. The problem is seen as a change point detection problem. Using operational calculus, which provides a convenient framework to handle such distributional derivatives, we characterize the time occurrence of a spike by an explicit formula. After spike detection we address the spike sorting problem. We developed a simple algorithm for a case when multi-channel recordings are available. The algorithm uses an iterative application of ICA and a deflation technique in two nested loops. In each iteration of the external loop, the spiking activity of one neuron is singled out and then deflated from the recordings. The internal loop implements a sequence of ICA and spike detection for removing the noise and all the spikes that are not coming from the targeted neuron. Finally, we discuss on properties of the neural code. We investigate whether the nature of the neural code is discrete or continuous. Moreover, if it is discrete, whether the elements of the code are drawn from a finite alphabet. We particularly address pulse-position coding scheme, making a link between communication theory and neural code.; Dans cette thèse, nous utilisons un ensemble d'outils avancés de mathématiques et de méthodes de traitement du signal pour aborder trois problèmes importants en neuroscience: la détection de potentiels d'action (spikes), leur tri, et le codage neuronal. A partir d'enregistrements invasifs de l'activité de neurones, nous avons d'abord abordé la question de la détection de spikes. Les instants d'occurrence des spikes apparaissent (comme des irrégularités) de manière explicite dans les dérivées distributionnelles du signal neuronal. Le problème est alors posé en termes de détection et estimation conjointe de ces instants d'occurrence. En utilisant le calcul opérationnel, qui fournit un cadre commode pour manipuler des Dirac, nous caractérisons le temps d'occurrence d'un spike par une formule explicite. Après la détection nous abordons le problème du tri de spikes. Nous avons développé un algorithme simple pour des enregistrements multi-canaux. Il utilise une application itérative de l'algorithme ICA et une technique de déflation dans deux boucles imbriquées. Dans chaque itération de la boucle externe, l'activité d'un neurone est ciblée, puis isolée et ensuite élaguée des enregistrements. La boucle interne met en œuvre une séquence d'applications de ICA et de détection afin de supprimer tout ce qui ne relève pas du neurone cible. Enfin, nous menons une discussion sur les propriétés du code neuronal. Nous examinons l'hypothèse d'un code reposant sur un alphabet fini. Cette hypothèse correspond à un schéma de modulation PPM qui est déjà suggéré par le codage temporel et la présence de jitter.

Comunicación presentada al "V VLSI Circuits and Systems" celebrado en Praga (República Checa) el 18 de Abril del 2011., This paper presents a power efficient architecture for a neural spike recording channel. The channel offers a selfcalibration operation mode and can be used both for signal tracking (to raw digitize the acquired neural waveform) and feature extraction (to build a PWL approximation of the spikes in order to reduce data bandwidth on the RF-link). The neural threshold voltage is adaptively calculated during the spike detection period using basic digital operations. The neural input signal is amplified and filtered using a LNA, reconfigurable Band-Pass Filter, followed by a fully reconfigurable 8-bit ADC. The key element is the ADC architecture. It is a binary search data converter with a SCimplementation. Due to its architecture, it can be programmed to work either as a PGA, S&H or ADC. In order to allow power saving, inactive blocks are powered off depending on the selected operation mode, ADC sampling frequency is reconfigured and bias current is dynamically adapted during the conversion. Due to the ADC low input capacitance, the power consumption of the input LNA can be decreased and the overall power consumption of the channel is low. The prototype was implemented using a CMOS 0.13um standard process, and it occupies 400um x 400um. Simulations from extracted layout show very promising results. The power consumption of the complete channel for the signal tracking operations is 2.8uW, and is increased to 3.0uW when the feature extraction operation is performed, one of the lowest reported., This work has been supported by the Spanish Ministry of Science and Innovation under grant TEC2009-08447, the Junta de Andalucía under grant TIC-02818 and the 2007-2013 FEDER Programme.

International audience; We present a study on wavelet detection methods of neuronal Action Potentials (APs). Our final goal is to implement the selected algorithms on custom integrated electronics for on-line processing of neural signals; therefore we take real-time computing as a hard specification and silicon area as a price to pay. Using simulated neural signals including APs, we characterize an efficient wavelet method for action potential extraction by evaluating its detection rate and its implementation cost. We compare software implementation for three methods: adaptive threshold, Discrete Wavelet Transform (DWT) and Stationary Wavelet Transform (SWT). We evaluate detection rate and implementation cost for detection functions dynamically comparing a signal with an adaptive threshold proportional to its standard deviation, where the signal is the raw neural signal, respectively: i) non processed; ii) processed by a Discrete Wavelet Transform (DWT); iii) processed by a Stationary Wavelet Transform (SWT). We also use different mother wavelets and test different data formats to set an optimal compromise between accuracy and silicon cost. Detection accuracy is evaluated together with false negative and false positive detections. Simulation results show that for on-line AP detection implemented on a configurable digital integrated circuit, APs underneath the noise level can be detected using SWT with a well-selected mother wavelet, combined to an adaptive threshold.

Signals from extracellular electrodes in neural systems record voltages resulting from activity in many neurons. Detecting action potentials (spikes) in a small number of specific (target) neurons is difficult because many neurons, both near and more distant, contribute to the signal at the electrode. We consider some nearby neurons as target neurons (providing a signal) and all the other contributions to the signal as noise. A new algorithm for spike detection has been developed: this applies a cepstrum of bispectrum (CoB) estimated inverse filter to provide blind equalization. This technique is based on higher order statistics, and seeks to find a sequence of event times or delta sequence. We show that the CoB-based technique can achieve a 98% hit rate on an extracellular signal containing three spike trains at up to 0 dB SNR. Threshold setting for this technique is discussed, and we show the application of the technique to some real signals. We compare performance with four established techniques and report that the CoB-based algorithm performs best.

International audience; Electroencephalography (EEG) occupies an important place for studying human brain activity in general, and epileptic processes in particular, with appropriate time resolution. Scalp EEG or intracerebral EEG signals recorded in patients with drug-resistant partial epilepsy convey important information about epileptogenic networks that must be localized and understood prior to subsequent therapeutic procedures. However, this information, often subtle, is 'hidden' in the signals. It is precisely the role of signal processing to extract this information and to put it into a 'coherent and interpretable picture' that can participate in the therapeutic strategy. Nowadays, the panel of available methods is very wide depending on the objectives such as, for instance, the detection of transient epileptiform events, the detection and/or prediction of seizures, the recognition and/or the classification of EEG patterns, the localization of epileptic neuronal sources, the characterization of neural synchrony, the determination of functional connectivity, among others. The intent of this paper is to focus on a specific category of methods providing relevant information about epileptogenic networks from the analysis of spatial properties of EEG signals in the time and frequency domain. These methods apply to either interictal or ictal recordings and share the common objective of localizing the subsets of brain structures involved in both types of paroxysmal activity. Most of these methods were developed by our group and are routinely used during pre-surgical evaluation. Examples are detailed. Results, as well as limitations of the methods, are also discussed.