Detecting temporal lobe seizures in ultra long-term subcutaneous EEG using algorithm-based data reduction.

OBJECTIVE: Ultra long-term monitoring with subcutaneous EEG (sqEEG) offers objective outpatient recording of electrographic seizures as an alternative to self-reported epileptic seizure diaries. This methodology requires an algorithm-based automatic seizure detection to indicate periods of potential seizure activity to reduce the time spent on visual review. The objective of this study was to evaluate the performance of a sqEEG-based automatic seizure detection algorithm. METHODS: A multicenter cohort of subjects using sqEEG were analyzed, including nine people with epilepsy (PWE) and 12 healthy subjects, recording a total of 965 days. The automatic seizure detections of a deep-neural-network algorithm were compared to annotations from three human experts. RESULTS: Data reduction ratios were 99.6% in PWE and 99.9% in the control group. The cross-PWE sensitivity was 86% (median 80%, range 69-100% when PWE were evaluated individually), and the corresponding median false detection rate was 2.4 detections per 24 hours (range: 2.0-13.0). CONCLUSIONS: Our findings demonstrated that step one in a sqEEG-based semi-automatic seizure detection/review process can be performed with high sensitivity and clinically applicable specificity. SIGNIFICANCE: Ultra long-term sqEEG bears the potential of improving objective seizure quantification.

Signal quality and power spectrum analysis of remote ultra long-term subcutaneous EEG.

OBJECTIVE: Ultra long-term subcutaneous electroencephalography (sqEEG) monitoring is a new modality with great potential for both health and disease, including epileptic seizure detection and forecasting. However, little is known about the long-term quality and consistency of the sqEEG signal, which is the objective of this study. METHODS: The largest multicenter cohort of sqEEG was analyzed, including 14 patients with epilepsy and 12 healthy subjects, implanted with a sqEEG device (24/7 EEG™ SubQ), and recorded from 23 to 230 days (median 42 days), with a median data capture rate of 75% (17.9 hours/day). Median power spectral density plots of each subject were examined for physiological peaks, including at diurnal and nocturnal periods. Long-term temporal trends in signal impedance and power spectral features were investigated with subject-specific linear regression models and group-level linear mixed-effects models. RESULTS: sqEEG spectrograms showed an approximate 1/f power distribution. Diurnal peaks in the alpha range (8-13Hz) and nocturnal peaks in the sigma range (12-16Hz) were seen in the majority of subjects. Signal impedances remained low, and frequency band powers were highly stable throughout the recording periods. SIGNIFICANCE: The spectral characteristics of minimally invasive, ultra long-term sqEEG are similar to scalp EEG, whereas the signal is highly stationary. Our findings reinforce the suitability of this system for chronic implantation on diverse clinical applications, from seizure detection and forecasting to brain-computer interfaces.

The Individual Ictal Fingerprint: Combining Movement Measures With Ultra Long-Term Subcutaneous EEG in People With Epilepsy.

Background: Epileptic seizures are caused by abnormal brain wave hypersynchronization leading to a range of signs and symptoms. Tools for detecting seizures in everyday life typically focus on cardiac rhythm, electrodermal activity, or movement (EMG, accelerometry); however, these modalities are not very effective for non-motor seizures. Ultra long-term subcutaneous EEG-devices can detect the electrographic changes that do not depend on clinical changes. Nonetheless, this also means that it is not possible to assess whether a seizure is clinical or subclinical based on an EEG signal alone. Therefore, we combine EEG and movement-related modalities in this work. We focus on whether it is possible to define an individual "multimodal ictal fingerprint" which can be exploited in different epilepsy management purposes. Methods: This study used ultra long-term data from an outpatient monitoring trial of persons with temporal lobe epilepsy obtained with a subcutaneous EEG recording system. Subcutaneous EEG, an EMG estimate and chest-mounted accelerometry were extracted from four persons showing more than 10 well-defined electrographic seizures each. Numerous features were computed from all three modalities. Based on these, the Gini impurity measure of a Random Forest classifier was used to select the most discriminative features for the ictal fingerprint. A total of 74 electrographic seizures were analyzed. Results: The optimal individual ictal fingerprints included features extracted from all three tested modalities: an acceleration component; the power of the estimated EMG activity; and the relative power in the delta (0.5-4 Hz), low theta (4-6 Hz), and high theta (6-8 Hz) bands of the subcutaneous EEG. Multimodal ictal fingerprints were established for all persons, clustering seizures within persons, while separating seizures across persons. Conclusion: The existence of multimodal ictal fingerprints illustrates the benefits of combining multiple modalities such as EEG, EMG, and accelerometry in future epilepsy management. Multimodal ictal fingerprints could be used by doctors to get a better understanding of the individual seizure semiology of people with epilepsy. Furthermore, the multimodal ictal fingerprint gives a better understanding of how seizures manifest simultaneously in different modalities. A knowledge that could be used to improve seizure acknowledgment when reviewing EEG without video.

Hypoglycemia-Associated EEG Changes Following Antecedent Hypoglycemia in Type 1 Diabetes Mellitus.

BACKGROUND: Recurrent hypoglycemia has been shown to blunt hypoglycemia symptom scores and counterregulatory hormonal responses during subsequent hypoglycemia. We therefore studied whether hypoglycemia-associated electroencephalogram (EEG) changes are affected by an antecedent episode of hypoglycemia. METHODS: Twenty-four patients with type 1 diabetes mellitus (10 with normal hypoglycemia awareness, 14 with hypoglycemia unawareness) were studied on 2 consecutive days by hyperinsulinemic glucose clamp at hypoglycemia (2.0-2.5 mmol/L) during a 1-h period. EEG was recorded, cognitive function assessed, and hypoglycemia symptom scores and counterregulatory hormonal responses were obtained. RESULTS: Twenty-one patients completed the study. Hypoglycemia-associated EEG changes were identified on both days with no differences in power or frequency distribution in the theta, alpha, or the combined theta-alpha band during hypoglycemia on the 2 days. Similar degree of cognitive dysfunction was also present during hypoglycemia on both days. When comparing the aware and unaware group, there were no differences in the hypoglycemia-associated EEG changes. There were very subtle differences in cognitive function between the two groups on day 2. The symptom response was higher in the aware group on both days, while only subtle differences were seen in the counterregulatory hormonal response. CONCLUSION: Antecedent hypoglycemia does not affect hypoglycemia-associated EEG changes in patients with type 1 diabetes mellitus.

Hypoglycemia-associated changes in the electroencephalogram in patients with type 1 diabetes and normal hypoglycemia awareness or unawareness.

Hypoglycemia is associated with increased activity in the low-frequency bands in the electroencephalogram (EEG). We investigated whether hypoglycemia awareness and unawareness are associated with different hypoglycemia-associated EEG changes in patients with type 1 diabetes. Twenty-four patients participated in the study: 10 with normal hypoglycemia awareness and 14 with hypoglycemia unawareness. The patients were studied at normoglycemia (5-6 mmol/L) and hypoglycemia (2.0-2.5 mmol/L), and during recovery (5-6 mmol/L) by hyperinsulinemic glucose clamp. During each 1-h period, EEG, cognitive function, and hypoglycemia symptom scores were recorded, and the counterregulatory hormonal response was measured. Quantitative EEG analysis showed that the absolute amplitude of the θ band and α-θ band up to doubled during hypoglycemia with no difference between the two groups. In the recovery period, the θ amplitude remained increased. Cognitive function declined equally during hypoglycemia in both groups and during recovery reaction time was still prolonged in a subset of tests. The aware group reported higher hypoglycemia symptom scores and had higher epinephrine and cortisol responses compared with the unaware group. In patients with type 1 diabetes, EEG changes and cognitive performance during hypoglycemia are not affected by awareness status during a single insulin-induced episode with hypoglycemia.

Automatic detection of childhood absence epilepsy seizures: toward a monitoring device.

Automatic detections of paroxysms in patients with childhood absence epilepsy have been neglected for several years. We acquire reliable detections using only a single-channel brainwave monitor, allowing for unobtrusive monitoring of antiepileptic drug effects. Ultimately we seek to obtain optimal long-term prognoses, balancing antiepileptic effects and side effects. The electroencephalographic appearance of paroxysms in childhood absence epilepsy is fairly homogeneous, making it feasible to develop patient-independent automatic detection. We implemented a state-of-the-art algorithm to investigate the performance of paroxysm detection. Using only a single scalp electroencephalogram channel from 20 patients with a total of 125 paroxysms >2 seconds, 97.2% of paroxysms could be detected with no false detections. This result leads us to recommend further investigations of tiny, one-channel electroencephalogram systems in an ambulatory setting.

Correlation between intra- and extracranial background EEG.

Scalp EEG is the most widely used modality to record the electrical signals of the brain. It is well known that the volume conduction of these brain waves through the brain, cerebrospinal fluid, skull and scalp reduces the spatial resolution and the signal amplitude. So far the volume conduction has primarily been investigated by realistic head models or interictal spike analysis. We have set up a novel and more realistic experiment that made it possible to compare the information in the intra- and extracranial EEG. We found that intracranial EEG channels contained correlated patterns when placed less than 30 mm apart, that intra- and extracranial channels were partly correlated when placed less than 40 mm apart, and that extracranial channels probably were correlated over larger distances. The underlying cortical area that influences the extracranial EEG is found to be up to 45 cm(2). This area is larger than previously reported.

Automatic seizure detection: going from sEEG to iEEG.

Several different algorithms have been proposed for automatic detection of epileptic seizure based on both scalp and intracranial electroencephalography (sEEG and iEEG). Which modality that renders the best result is hard to assess though. From 16 patients with focal epilepsy, at least 24 hours of ictal and non-ictal iEEG were obtained. Characteristics of the seizures are represented by use of wavelet transformation (WT) features and classified by a support vector machine. When implementing a method used for sEEG on iEEG data, a great improvement in performance was obtained when the high frequency containing lower levels in the WT were included in the analysis. We were able to obtain a sensitivity of 96.4% and a false detection rate (FDR) of 0.20/h. In general, when implementing an automatic seizure detection algorithm made for sEEG on iEEG, great improvement can be obtained if a frequency band widening of the feature extraction is performed. This means that algorithms for sEEG should not be discarded for use on iEEG - they should be properly adjusted as exemplified in this paper.