Automatic seizure detection: improvements and evaluation.

Improvements to an existing automatic seizure detection program are described. They are aimed at taking into account a larger temporal context and thus improving the specificity of the detections. Results were evaluated on 293 recordings from 49 patients, totaling 5303 h of 16-channel recording. They showed that 24% of the 244 seizures recorded were missed by the automatic detection; in 41% of the seizures, the patient alarm was not pressed but the computer made detections. The false detection rate was of the order of 1 false detection per hour of recording. Conclusions are: (1) automatic seizure detection must be used in conjunction with a patient alarm button since some seizures, having poorly defined EEG activity, are not detected; (2) the automatic detection allowed capture of many seizures, clinical and subclinical, for which the alarm was not pressed; (3) the low false detection rate indicates that lower detection threshold could be used, yielding better seizure detection.     

Neuroanatomical correlates of biological motion detection

Biological motion detection is both commonplace and important, but there is great inter-individual variability in this ability, the neural basis of which is currently unknown. Here we examined whether the behavioral variability in biological motion detection is reflected in brain anatomy. Perceptual thresholds for detection of biological motion and control conditions (non-biological object motion detection and motion coherence) were determined in a group of healthy human adults (n=31) together with structural magnetic resonance images of the brain. Voxel based morphometry analyzes revealed that gray matter volumes of left posterior superior temporal sulcus (pSTS) and left ventral premotor cortex (vPMC) significantly predicted individual differences in biological motion detection, but showed no significant relationship with performance on the control tasks. Our study reveals a neural basis associated with the inter-individual variability in biological motion detection, reliably linking the neuroanatomical structure of left pSTS and vPMC with biological motion detection performance.     

The effect of psychological stress and relaxation on interoceptive accuracy: Implications for symptom perception

OBJECTIVES: The goals of the current study were to investigate: (i) how the manipulation of psychophysiological state (stress vs. relaxation) would influence heartbeat detection performance in a laboratory environment and (ii) whether interoceptive accuracy had a relationship with symptom reporting. METHOD: Forty participants (20 males) performed a stressor (a demanding mental arithmetic task) and a relaxation exercise during two counterbalanced sessions, both of which included baseline (control) conditions. Performance of both tasks was interspersed with a heartbeat detection task, i.e., a two-choice Whitehead paradigm. Data were collected from subjective mood scales as well as the electrocardiogram. RESULTS: Both stress and relaxation conditions had the anticipated influence on subjective mood. There was no effect of stress or relaxation on heartbeat detection accuracy for male participants. However, the heartbeat detection accuracy of female participants showed a significant decline during the stressor condition. There was evidence that lower mean heart rate tended to improve heartbeat detection performance. A regression analysis revealed that two traits from the Body Perception Questionnaire (autonomic reactivity and body awareness) predicted heartbeat detection accuracy but not in the expected direction. CONCLUSIONS: The study provided evidence of a gender-specific decrement of heartbeat detection accuracy due to a laboratory stressor. However, the relevance of this finding for health psychology may be limited, as interoceptive accuracy had no significant relationship with symptom reporting.     

Detection of Brain Activation Signal from Functional Magnetic Resonance Imaging Data

An image–processing strategy for functional magnetic resonance imaging (fMRI) data sets consisting of sequential images of the same slice of brain tissue is considered. An algorithm of detection based on the likelihood–ratio test and the noise properties in fMRI is introduced. Since the data have a poor signal–to–noise ratio, and in order to make detection reliable, the algorithm is organized in two steps: (1) pixel detection, which detects all pixels having significant changes, thus building regions of interest (ROIs), and (2) region detection, which selects the most likely activated region from obtained ROIs. The detection method is applied to experimental fMRI data from the motor cortex and compared with the cross–correlation method and Student's t test commonly applied by others. The results obtained using the likelihood–ratio test show improvement in the detection of activated regions.     

Detection of focal epileptiform events in the EEG by spatio-temporal dipole clustering

OBJECTIVE: Methods for the detection of epileptiform events can be broadly divided into two main categories: temporal detection methods that exploit the EEG's temporal characteristics, and spatial detection methods that base detection on the results of an implicit or explicit source analysis. We describe how the framework of a spatial detection method was extended to improve its performance by including temporal information. This results in a method that provides (i) automated localization of an epileptogenic focus and (ii) detection of focal epileptiform events in an EEG recording. For the detection, only one threshold value needs to be set. METHODS: The method comprises five consecutive steps: (1) dipole source analysis in a moving window, (2) automatic selection of focal brain activity, (3) dipole clustering to arrive at the identification of the epileptiform cluster, (4) derivation of a spatio-temporal template of the epileptiform activity, and (5) template matching. Routine EEG recordings from eight paediatric patients with focal epilepsy were labelled independently by two experts. The method was evaluated in terms of (i) ability to identify the epileptic focus, (ii) validity of the derived template, and (iii) detection performance. The clustering performance was evaluated using a leave-one-out cross validation. Detection performance was evaluated using Precision-Recall curves and compared to the performance of two temporal (mimetic and wavelet based) and one spatial (dipole analysis based) detection methods. RESULTS: The method succeeded in identifying the epileptogenic focus in seven of the eight recordings. For these recordings, the mean distance between the epileptic focus estimated by the method and the region indicated by the labelling of the experts was 8mm. Except for two EEG recordings where the dipole clustering step failed, the derived template corresponded to the epileptiform activity marked by the experts. Over the eight EEGs, the method showed a mean sensitivity and selectivity of 92 and 77%, respectively. CONCLUSIONS: The method allows automated localization of the epileptogenic focus and shows good agreement with the region indicated by the labelling of the experts. If the dipole clustering step is successful, the method allows a detection of the focal epileptiform events, and gave a detection performance comparable or better to that of the other methods. SIGNIFICANCE: The identification and quantification of epileptiform events is of considerable importance in the diagnosis of epilepsy. Our method allows the automatic identification of the epileptic focus, which is of value in epilepsy surgery. The method can also be used as an offline exploration tool for focal EEG activity, displaying the dipole clusters and corresponding time series.     

Intracerebrally recorded high frequency oscillations: Simple visual assessment versus automated detection

Objective

We compared the possible contribution (in the detection of seizure onset zone – SOZ) of simple visual assessment of intracerebrally recorded high-frequency oscillations (HFO) with standard automated detection.

Methods

We analyzed stereo-EEG (SEEG) recordings from 20 patients with medically intractable partial seizures (10 temporal/10 extratemporal). Independently using simple visual assessment and automated detection of HFO, we identified the depth electrode contacts with maximum occurrences of ripples (R) and fast ripples (FR). The SOZ was determined by independent visual identification in standard SEEG recordings, and the congruence of results from visual versus automated HFO detection was compared.

Results

Automated detection of HFO correctly identified the SOZ in 14 (R)/10 (FR) out of 20 subjects; a simple visual assessment of SEEG recordings in the appropriate frequency ranges correctly identified the SOZ in 13 (R)/9 (FR) subjects.

Conclusions

Simple visual assessment of SEEG traces and standard automated detection of HFO seem to contribute comparably to the identification of the SOZ in patients with focal epilepsies. When using macroelectrodes in neocortical extratemporal epilepsies, the SOZ might be better determined by the ripple range.

Significance

Standard automated detection of HFO enables the evaluation of HFO characteristics in whole data. This detection allows general purpose and objective evaluation, without any bias from the neurophysiologist’s experiences and practice.     

MRI与三维容积超声在新生儿颅内出血诊断中的临床应用

目的探讨MRI与三维容积超声在新生儿颅内出血诊断中的临床价值。方法选择我院收治的新生儿颅内出血患儿80例,均实施MRI与三维容积超声检查。结果 80例患儿中,MRI颅内出血检出率70.00%(56/80),三维容积超声为53.75%(43/80),差异有统计学意义(P0.05);三维容积超声检出出血量(4.96±0.87)mL,MRI为(4.63±0.47)mL,差异有统计学意义(P0.05)。结论 MRI与三维容积超声在新生儿颅内出血诊断中各有利弊,三维容积超声有利于检出新生儿颅内出血量,MRI在新生儿颅内出血检出率方面更有优势,临床可根据患儿的自身病情选择合理的检查方式。     

Diazepam-induced changes in signal detection performance. A comparison with the effects on the Critical Flicker-Fusion Frequency and the Digit Symbol Substitution Test

The effects of 10 mg diazepam on signal detection theory measures (stimulus sensitivity, response bias) and reaction times were studied in a 1-hour visual signal detection task with high and low signal probability, and on performance in two short-duration tasks: Critical Flicker-Fusion Frequency (CFF) and the Digit Symbol Substitution Test (DSST). 12 healthy volunteers participated in this placebo-controlled, double-blind cross-over study. Diazepam affected the stimulus sensitivity and the reaction times of hits in the signal detection task. DSST performance was also impaired while CFF did not change after diazepam treatment. No relationship between serum diazepam concentration and change in task performance was found. It is concluded that diazepam affects signal detection performance, independent of signal probability. A short-duration task like the DSST is as sensitive to the effects of diazepam as the (long-duration) signal detection task.     

Global motion detection is impaired in cats deprived early of pattern vision

We investigated global motion detection in binocularly deprived cats (BD cats) and control cats (C cats). The cats were trained in the two-choice free running apparatus for a food reward. The positive stimulus was a moving random-dot pattern with all dots moving in one direction, the negative stimulus was the same random-dot pattern but stationary. The BD cats were severely impaired in detection of global motion stimulus as compared with the C cats. In contrast, their level of performance in a simple relative motion detection task (one square) did not differ from that in the C cats. However, in more complex relative motion detection task (two squares) the performance of the BD cats was impaired. The deficit in the detection of global motion in BD cats may be due to impairments of their Y-pathway.     

Predictors of atrial fibrillation detection and features of recurrent strokes in patients after cryptogenic stroke

Background and purposeUse of implantable cardiac monitors (ICMs) has increased diagnosis of atrial fibrillation (AF) in cryptogenic stroke (CS) patients. Identifying AF predictors may enhance the yield of AF detection. Recurrent strokes after CS are not well described. We aimed to assess the predictors for AF detection and the characteristics of recurrent strokes in patients after CS.MethodsWe reviewed electronic medical records of CS patients who were admitted between February 2014 and September 2017 and underwent ICM placement with minimum one-year follow-up. Patient demographics, stroke characteristics, pre-defined risk factors as well as recurrent strokes were compared between patients with and without AF detection.Results389 patients with median follow-up of 548 days were studied. AF was detected in 102 patients (26.2%). Age (per decade increase, OR 2.10, CI 1.64–2.68, with vs. without AF) and left atrium diameter (per 5 mm increase, OR 1.91, CI 1.33–2.74) were identified as AF predictors. Intracranial large vessel stenosis >50% irrelevant to the index strokes was associated with AF detection within 30 days (OR 0.24, CI 0.09–0.69, >30 vs. <30 days). Recurrent strokes occurred in 14% patients with median follow-up about 2.5 years. Topography of these strokes resembled embolic pattern and was comparable between patients with and without AF. Among recurrent strokes in patients with AF, the median time to AF detection was much shorter (90 vs. 251 days), and the median time to first stroke recurrence was much longer (422 vs. 76 days) in patients whose strokes recurred after AF detection than those before AF detection.ConclusionsOlder age and enlarged left atrium are predictors for AF detection in CS patients. Intracranial atherosclerosis is more prevalent in patients with early AF detection within 30 days. Recurrent strokes follow the embolic pattern, and early AF detection could delay the stroke recurrence.     

Combination of EEG and ECG for improved automatic neonatal seizure detection.

OBJECTIVE: Neonatal seizures are the most common central nervous system disorder in newborn infants. A system that could automatically detect the presence of seizures in neonates would be a significant advance facilitating timely medical intervention. METHODS: A novel method is proposed for the robust detection of neonatal seizures through the combination of simultaneously-recorded electroencephalogram (EEG) and electrocardiogram (ECG). A patient-specific and a patient-independent system are considered, employing statistical classifier models. RESULTS: Results for the signals combined are compared to results for each signal individually. For the patient-specific system, 617 of 633 (97.52%) expert-labelled seizures were correctly detected with a false detection rate of 13.18%. For the patient-independent system, 516 of 633 (81.44%) expert-labelled seizures were correctly detected with a false detection rate of 28.57%. CONCLUSIONS: A novel algorithm for neonatal seizure detection is proposed. The combination of an ECG-based classifier system with a novel multi-channel EEG-based classifier system has led to improved seizure detection performance. The algorithm was evaluated using a large data-set containing ECG and multi-channel EEG of realistic duration and quality. SIGNIFICANCE: Analysis of simultaneously-recorded EEG and ECG represents a new approach in seizure detection research and the detection performance of the proposed system is a significant improvement on previous reported results for automated neonatal seizure detection.     

Optimized deep neural network architecture for robust detection of epileptic seizures using EEG signals

ObjectiveAutomatic detection of epileptic seizures based on deep learning methods received much attention last year. However, the potential of deep neural networks in seizure detection has not been fully exploited in terms of the optimal design of the model architecture and the detection power of the time-series brain data. In this work, a deep neural network architecture is introduced to learn the temporal dependencies in Electroencephalogram (EEG) data for robust detection of epileptic seizures.MethodsA deep Long Short-Term Memory (LSTM) network is first used to learn the high-level representations of different EEG patterns. Then, a Fully Connected (FC) layer is adopted to extract the most robust EEG features relevant to epileptic seizures. Finally, these features are supplied to a softmax layer to output predicted labels.ResultsThe results on a benchmark clinical dataset reveal the prevalence of the proposed approach over the baseline techniques; achieving 100% classification accuracy, 100% sensitivity, and 100% specificity. Our approach is additionally shown to be robust in noisy and real-life conditions. It maintains high detection performance in the existence of common EEG artifacts (muscle activities and eye movement) as well as background noise.ConclusionsWe demonstrate the clinical feasibility of our seizure detection approach achieving superior performance over the cutting-edge techniques in terms of seizure detection performance and robustness.SignificanceOur seizure detection approach can contribute to accurate and robust detection of epileptic seizures in ideal and real-life situations.     

A multistage knowledge-based system for EEG seizure detection in newborn infants

OBJECTIVE: Automatic seizure detection has attracted attention as a method to obtain valuable information concerning the duration, timing, and frequency of seizures. Methods currently used to detect EEG seizures in adults show high false detection rates in neonates because they lack information about specific age-dependent features of normal and pathological EEG and artifacts. This paper describes a novel multistage knowledge-based seizure detection system for newborn infants to identify and classify normal and pathological newborn EEGs as well as seizures with a reduced false detection rate. METHODS: We developed the system in a way to make comprehensive use of spatial and temporal contextual information obtained from multichannel EEGs. The system development consists of six major stages: (i) EEG data collection and bandpass filtering; (ii) automatic artifact detection; (iii) feature extraction from segments of non-seizure and seizure activities; (iv) feature selection via the relevance and redundancy analysis; (v) EEG classification and pattern recognition using a trained multilayer back-propagation neural network; and (v) knowledge-based decision-making to examine each of possible EEG patterns from a multi-channel perspective. The system was developed and tested with the EEG recordings of 10 newborns aged between 39 and 42 weeks. RESULTS: The overall sensitivity, selectivity, and average detection rate of the system were 74%, 70.1%, and 79.7%, respectively. The average false detection of 1.55/h was also achieved by the system with a feature reduction up to 80%. CONCLUSIONS: The expert rule-based decision-making subsystem accompanying the classifier helped to reduce the false detection rate, reject a wide variety of artifacts, and discriminate various patterns of EEG. SIGNIFICANCE: This paper may serve as a guide for the selection of discriminative features to improve the accuracy of conventional seizure detection systems for routine clinical EEG interpretation and brain activity monitoring in newborns especially those hospitalized in the neonatal intensive care units.     

Assessment of in vitro-generated platelet microparticles using a modified flow cytometric strategy.

Quantification of platelet microparticles (PMPs) may be a useful marker for the detection of in vivo platelet activation. Optimisation of flow cytometric methods for detection and quantification of PMPs has not been systemically evaluated. This study reports the optimisation of flow cytometric procedures for the detection of PMPs, the determination of limits of size detection using microbeads, and the characterisation of PMP generation by in vitro activation of platelets using collagen and adenosine 5' diphosphate (ADP). Fluorescent and plain microbeads proved useful for defining the limits of the flow cytometer in detecting PMPs. A systematic calibration of the forward scatter (FS) threshold parameter (size) of the flow cytometer using microbeads allowed for the detection of very small particles (down to 0.1 microm diameter). PMPs generated in vitro using ADP and collagen were reliably detected by flow cytometry using monoclonal antibodies (MAb) directed towards platelet surface membrane glycoproteins (Gp). The PMP events were detected in the FS low (i.e., small size events) and fluorescence (FL) high (i.e., platelet Gp MAb-labelled events) region. PMPs of different size profiles were observed for each of the agonists. Flow cytometry can be used as a tool in the assessment of PMPs. As detection of particles of this type is at the limit of resolution of flow cytometers, careful attention is required with the choice of platelet-specific MAb, isotype control, and optimisation of procedure setup and performance.     

Physicians' low detection rates of alcohol dependence or abuse: a matter of methodological shortcomings?

Previous research may have underestimated physicians' detection rates of alcohol dependence or abuse because case findings have been based on screening questionnaires instead of using in-depth diagnostic criteria and detection rates have been assessed by analyzing patient records instead of directly interviewing the physician. To test this hypothesis, consecutive patients of a general hospital (N=436) and of 12 randomly selected general practices (N=929) were examined. A two-step diagnostic procedure included screening questionnaires and a diagnostic interview (SCAN). The analysis compares detection rates based on methods used in previous studies to data using more precise methods. Physicians' detection rates ranged from 37.0% to 88.9% in the general hospital and from 11.1% to 74.7% in general practices depending on methods used. The physicians' detection rates could be improved by 10% (general hospital) and 20% (general practice) through the additional use of a screening questionnaire. Of those patients assessed by the physicians as problem drinkers in the general hospital, 13.9% were referred to an addiction consultation-liaison service. Data reveal that physicians' abilities to detect problem drinkers have been underestimated. Routine screening procedures could play a major role in improving detection rates and reminding the physician to intervene.     

Tactile direction discrimination and vibration detection in diabetic neuropathy

Löken LS, Lundblad LC, Elam M, Olausson HW. Tactile direction discrimination and vibration detection in diabetic neuropathy. Acta Neurol Scand: 2010: 121: 302–308.
© 2009 The Authors Journal compilation © 2009 Blackwell Munksgaard. Objective – To evaluate the clinical usefulness of quantitative testing of tactile direction discrimination (TDD) in patients with diabetic neuropathy. Materials and methods – TDD and vibration detection were examined on the dorsum of the feet in 43 patients with type 1 diabetes mellitus and clinical signs and symptoms indicating mild neuropathy, and abnormal results for neurography, temperature detection, or heart rate variability. Test–retest examination of TDD was performed in nine of the patients. Results – Twenty‐six of the patients had abnormal TDD (sensitivity 0.60) and 20 had abnormal vibration detection (sensitivity 0.46). Ten of the patients had abnormal TDD and normal vibration detection. Four of the patients had abnormal vibration detection and normal TDD. Test–retest examination of TDD showed a high degree of reproducibility (r = 0.87). Conclusion – TDD seems more useful than vibration detection in examination of diabetic neuropathy.     

Image level detection of large vessel occlusion on 4D-CTA perfusion data using deep learning in acute stroke

ObjectivesAutomated image-level detection of large vessel occlusions (LVO) could expedite patient triage for mechanical thrombectomy. A few studies have previously attempted LVO detection using artificial intelligence (AI) on CT angiography (CTA) images. To our knowledge this is the first study to detect LVO existence and location on raw 4D-CTA/ CT perfusion (CTP) images using neural network (NN) models.Materials and MethodsRetrospective study using data from a level-I stroke center was performed. A total of 306 (187 with LVO, and 119 without) patients were evaluated. Image pre-processing included co-registration, normalization and skull stripping. Five consecutive time-points for each patient were selected to provide variable contrast density in data. Additional data augmentation included rotation and horizonal image flipping. Our model architecture consisted of two neural networks, first for classification (based on hemispheric asymmetry), followed by second model for exact site of LVO detection. Only cases deemed positive by the classification model were routed to the detection model, thereby reducing false positives and improving specificity. The results were compared with expert annotated LVO detection.ResultsUsing a 80:20 split for training and validation, the combination of both classification and detection model achieved a sensitivity of 86.5%, a specificity of 89.5%, and an accuracy of 87.5%. A 5-fold cross-validation using the entire data achieved a mean sensitivity of 82.7%, a specificity of 89.8%, and an accuracy of 85.5% and a mean AUC of 0.89 (95% CI: 0.85-0.93).ConclusionOur findings suggest that accurate image-level LVO detection is feasible on CTP raw images.     

Recognition of emotional prosody and verbal components of spoken language: an fMRI study

This study examined the neural areas involved in the recognition of both emotional prosody and phonemic components of words expressed in spoken language using echo-planar, functional magnetic resonance imaging (fMRI). Ten right-handed males were asked to discriminate words based on either expressed emotional tone (angry, happy, sad, or neutral) or phonemic characteristics, specifically, initial consonant sound (bower, dower, power, or tower). Significant bilateral activity was observed in the detection of both emotional and verbal aspects of language when compared to baseline activity. We found that the detection of emotion compared with verbal detection resulted in significant activity in the right inferior frontal lobe. Conversely, the detection of verbal stimuli compared with the detection of emotion activated left inferior frontal lobe regions most significantly. Specific analysis of the anterior auditory cortex revealed increased right hemisphere activity during the detection of emotion compared to activity during verbal detection. These findings illustrate bilateral involvement in the detection of emotion in language while concomitantly showing significantly lateralized activity in both emotional and verbal detection, in both the temporal and frontal lobes.