Combining machine learning models for the automatic detection of EEG arousals

Exploring Convolutional Neural Network Architectures for EEG Feature Extraction

The main purpose of this paper is to provide information on how to create a convolutional neural network (CNN) for extracting features from EEG signals. Our task was to understand the primary aspects of creating and fine-tuning CNNs for various application scenarios. We considered the characteristics of EEG signals, coupled with an exploration of various signal processing and data preparation techniques. These techniques include noise reduction, filtering, encoding, decoding, and dimension reduction, among others. In addition, we conduct an in-depth analysis of well-known CNN architectures, categorizing them into four distinct groups: standard implementation, recurrent convolutional, decoder architecture, and combined architecture. This paper further offers a comprehensive evaluation of these architectures, covering accuracy metrics, hyperparameters, and an appendix that contains a table outlining the parameters of commonly used CNN architectures for feature extraction from EEG signals.

Multi-task learning for arousal and sleep stage detection using fully convolutional networks

Objective.Sleep is a critical physiological process that plays a vital role in maintaining physical and mental health. Accurate detection of arousals and sleep stages is essential for the diagnosis of sleep disorders, as frequent and excessive occurrences of arousals disrupt sleep stage patterns and lead to poor sleep quality, negatively impacting physical and mental health. Polysomnography is a traditional method for arousal and sleep stage detection that is time-consuming and prone to high variability among experts.Approach. In this paper, we propose a novel multi-task learning approach for arousal and sleep stage detection using fully convolutional neural networks. Our model, FullSleepNet, accepts a full-night single-channel EEG signal as input and produces segmentation masks for arousal and sleep stage labels. FullSleepNet comprises four modules: a convolutional module to extract local features, a recurrent module to capture long-range dependencies, an attention mechanism to focus on relevant parts of the input, and a segmentation module to output final predictions.Main results.By unifying the two interrelated tasks as segmentation problems and employing a multi-task learning approach, FullSleepNet achieves state-of-the-art performance for arousal detection with an area under the precision-recall curve of 0.70 on Sleep Heart Health Study and Multi-Ethnic Study of Atherosclerosis datasets. For sleep stage classification, FullSleepNet obtains comparable performance on both datasets, achieving an accuracy of 0.88 and an F1-score of 0.80 on the former and an accuracy of 0.83 and an F1-score of 0.76 on the latter.Significance. Our results demonstrate that FullSleepNet offers improved practicality, efficiency, and accuracy for the detection of arousal and classification of sleep stages using raw EEG signals as input.

Machine learning approaches for predicting sleep arousal response based on heart rate variability, oxygen saturation, and body profiles

Objective

Obstructive sleep apnea is a global health concern, and several tools have been developed to screen its severity. However, most tools focus on respiratory events instead of sleep arousal, which can also affect sleep efficiency. This study employed easy-to-measure parameters-namely heart rate variability, oxygen saturation, and body profiles-to predict arousal occurrence.

Methods

Body profiles and polysomnography recordings were collected from 659 patients. Continuous heart rate variability and oximetry measurements were performed and then labeled based on the presence of sleep arousal. The dataset, comprising five body profiles, mean heart rate, six heart rate variability, and five oximetry variables, was then split into 80% training/validation and 20% testing datasets. Eight machine learning approaches were employed. The model with the highest accuracy, area under the receiver operating characteristic curve, and area under the precision recall curve values in the training/validation dataset was applied to the testing dataset and to determine feature importance.

Results

InceptionTime, which exhibited superior performance in predicting sleep arousal in the training dataset, was used to classify the testing dataset and explore feature importance. In the testing dataset, InceptionTime achieved an accuracy of 76.21%, an area under the receiver operating characteristic curve of 84.33%, and an area under the precision recall curve of 86.28%. The standard deviations of time intervals between successive normal heartbeats and the square roots of the means of the squares of successive differences between normal heartbeats were predominant predictors of arousal occurrence.

Conclusions

The established models can be considered for screening sleep arousal occurrence or integrated in wearable devices for home-based sleep examination.

Entropy-Based Emotion Recognition from Multichannel EEG Signals Using Artificial Neural Network

Humans experience a variety of emotions throughout the course of their daily lives, including happiness, sadness, and rage. As a result, an effective emotion identification system is essential for electroencephalography (EEG) data to accurately reflect emotion in real-time. Although recent studies on this problem can provide acceptable performance measures, it is still not adequate for the implementation of a complete emotion recognition system. In this research work, we propose a new approach for an emotion recognition system, using multichannel EEG calculation with our developed entropy known as multivariate multiscale modified-distribution entropy (MM-mDistEn) which is combined with a model based on an artificial neural network (ANN) to attain a better outcome over existing methods. The proposed system has been tested with two different datasets and achieved better accuracy than existing methods. For the GAMEEMO dataset, we achieved an average accuracy ± standard deviation of 95.73% ± 0.67 for valence and 96.78% ± 0.25 for arousal. Moreover, the average accuracy percentage for the DEAP dataset reached 92.57% ± 1.51 in valence and 80.23% ± 1.83 in arousal.

DeepSleep convolutional neural network allows accurate and fast detection of sleep arousal

Sleep arousals are transient periods of wakefulness punctuated into sleep. Excessive sleep arousals are associated with symptoms such as sympathetic activation, non-restorative sleep, and daytime sleepiness. Currently, sleep arousals are mainly annotated by human experts through looking at 30-second epochs (recorded pages) manually, which requires considerable time and effort. Here we present a deep learning approach for automatically segmenting sleep arousal regions based on polysomnographic recordings. Leveraging a specific architecture that 'translates' input polysomnographic signals to sleep arousal labels, this algorithm ranked first in the "You Snooze, You Win" PhysioNet Challenge. We created an augmentation strategy by randomly swapping similar physiological channels, which notably improved the prediction accuracy. Our algorithm enables fast and accurate delineation of sleep arousal events at the speed of 10 seconds per sleep recording. This computational tool would greatly empower the scoring process in clinical settings and accelerate studies on the impact of arousals.

Automatic detection of cortical arousals in sleep and their contribution to daytime sleepiness

OBJECTIVE

Significant interscorer variability is found in manual scoring of arousals in polysomnographic recordings (PSGs). We propose a fully automatic method, the Multimodal Arousal Detector (MAD), for detecting arousals.

METHODS

A deep neural network was trained on 2,889 PSGs to detect cortical arousals and wakefulness in 1-second intervals. Furthermore, the relationship between MAD-predicted labels on PSGs and next day mean sleep latency (MSL) on a multiple sleep latency test (MSLT), a reflection of daytime sleepiness, was analyzed in 1447 MSLT instances in 873 subjects.

RESULTS

In a dataset of 1,026 PSGs, the MAD achieved an F1 score of 0.76 for arousal detection, while wakefulness was predicted with an accuracy of 0.95. In 60 PSGs scored by nine expert technicians, the MAD performed comparable to four and significantly outperformed five expert technicians for arousal detection. After controlling for known covariates, a doubling of the arousal index was associated with an average decrease in MSL of 40 seconds (p = 0.0075).

CONCLUSIONS

The MAD performed better or comparable to human expert scorers. The MAD-predicted arousals were shown to be significant predictors of MSL.

SIGNIFICANCE

This study validates a fully automatic method for scoring arousals in PSGs.

Trend Statistics Network and Channel invariant EEG Network for sleep arousal study

Sleep is a very important part of life. Lack of sleep or sleep disorder can cause a negative impact on day to day life and can have long term serious consequences. In this work, we propose an end-to-end trainable neural network for automated sleep arousal scoring. The network consists of two main parts. Firstly, a trend statistics network computes the moving average of the filtered signals at different scales. Secondly, we propose a channel invariant EEG network to detect the arousals in any Electroencephalography (EEG) channel. Finally, we combine the features from various channels through a convolution network and a bi-directional long short-term memory to predict the probability of arousal. Further, we propose an objective function that uses only respiratory effort related arousal (RERA) and non-arousal regions to optimize the network. We also propose a method to estimate the respiratory disturbance index (RDI) from the probability predicted by the network. Evaluation on Physionet Challenge 2018 database shows that the proposed method detects RERA with mean area under the precision-recall curve (AUPRC) of 0.50 in a 10-fold cross validation setup. The mean absolute error of RDI prediction is 6.11, while a two-class RDI severity prediction yields a specificity of 75% and a sensitivity of 83%.

An efficient automatic arousals detection algorithm in single channel EEG

BACKGROUND AND OBJECTIVE

Electroencephalographic arousal is a transient waveform that instantaneously happens in sleep as an inherent component. It has distinctive amplitude and frequency features. However, it is visually difficult to distinguish arousal from the background of the electroencephalogram. This visual scoring is important for brain researches, sleep studies, sleep stage scorings and assessment of sleep disorders. The scoring process is a time-consuming and difficult clinical procedure which is evaluated by sleep experts. It may also have subjective consequences due to the variability of personal expertise of physicians. Conversely, this scoring process can be significantly accelerated with computer-aided automated algorithms. Moreover, reproducible and objective results can be obtained. In this work, we propose a novel algorithm for the automatic detection of electroencephalographic arousals in sleep polysomnographic recordings.

METHODS

The approach uses a well-known time-frequency localization method, the continuous wavelet transform, to identify relevant arousal patterns. Special emphasis was carried out to produce a robust, reliable, fast and artifact tolerant algorithm. In the first part, the electroencephalographic scalogram, the squared magnitude of the continuous wavelet transform, was obtained. The mean and variance of the scalogram coefficients were determined as novel features. Support vector machine was applied as a classifier. Half of the recordings were used for training with five-fold cross-validation and a high accuracy training rate was obtained. Then, the rest of the recordings were used for testing.

RESULTS

As a result, the overall sensitivity, specificity, accuracy, and positive predictive value of the algorithm are 94.67%, 99.33%, 98.2%, and 97.93%, respectively.

CONCLUSION

In this paper, we have shown that the electroencephalographic arousal pattern can be characterized by the scalogram in the wavelet domain. The proposed algorithm works with high accuracy, reproducibility and gives objective results without case-specific sensitivity.

Iterative expert-in-the-loop classification of sleep PSG recordings using a hierarchical clustering

BACKGROUND

The classification of sleep signals is a subjective and time consuming task. A large number of automatic classifiers have been published in the past decade but a sleep community has no strong confidence to use them in clinical practice and still remains using a standard manual scoring according standardized rules.

NEW METHOD

We developed a semi-supervised data-driven approach for objective and efficient evaluation of polysomnographic (PSG) data. The proposed algorithm finds a representative set of signal segments that are subsequently scored by a sleep neurologist. The remaining part of the recording is then automatically classified using these templates.

RESULTS

The method was evaluated on 36 PSG recordings (18 chronic insomniacs, 18 healthy controls). We show a faster and objective evaluation of PSG data compared to the manual scoring that is over-performing automated classifiers (accuracy increases ∼14%). The classification results are comparable on both datasets.

COMPARISON WITH EXISTING METHOD(S)

The methodology that we propose has not yet been published in the area of sleep PSG data processing. The performance of our method is comparable to various published automated approaches (a typical published classification accuracy is ∼75-95%). The method allows the evaluation of PSG recordings in more general terms and across different recording devices and standards.

CONCLUSIONS

The proposed solution is not based on a single-purpose rules or heuristics and training model is not trained on other patient's sleep recordings. The method is applicable to wide range of similar tasks and various types of physiological signals.

Large-scale validation of an automatic EEG arousal detection algorithm using different heterogeneous databases

OBJECTIVE

To assess the validity of an automatic EEG arousal detection algorithm using large patient samples and different heterogeneous databases.

METHODS

Automatic scorings were confronted with results from human expert scorers on a total of 2768 full-night PSG recordings obtained from two different databases. Of them, 472 recordings were obtained during a clinical routine at our sleep center and were subdivided into two subgroups of 220 (HMC-S) and 252 (HMC-M) recordings each, according to the procedure followed by the clinical expert during the visual review (semi-automatic or purely manual, respectively). In addition, 2296 recordings from the public SHHS-2 database were evaluated against the respective manual expert scorings.

RESULTS

Event-by-event epoch-based validation resulted in an overall Cohen's kappa agreement of κ = 0.600 (HMC-S), 0.559 (HMC-M), and 0.573 (SHHS2). Estimated inter-scorer variability on the datasets was, respectively, κ = 0.594, 0.561 and 0.543. Analyses of the corresponding Arousal Index scores showed associated automatic-human repeatability indices ranges of 0.693-0.771 (HMC-S), 0.646-0.791 (HMC-M), and 0.759-0.791 (SHHS2).

CONCLUSIONS

Large-scale validation of our automatic EEG arousal detector on different databases has shown robust performance and good generalization results comparable to the expected levels of human agreement. Special emphasis was put on reproducibility of the results; implementation of our method has been made available online as open source code.

A simple and robust method for the automatic scoring of EEG arousals in polysomnographic recordings

BACKGROUND

Clinical diagnosis of sleep disorders relies on the polysomnographic test to examine the neurophysiological markers of the sleep process. In this test, the recording of the electroencephalographic activity and the submental electromyogram is the source of the analysis for the detection of electroencephalographic arousals. The identification of these events is important for the evaluation of the sleep continuity because they cause the fragmentation of the normal sleep process. This work proposes a new technique for the automatic detection of arousals in polysomnographic recordings, presenting a non-computationally complex method with the idea of providing an easy integration with other algorithms.

METHODS

The proposed algorithm combines different well-known signal analysis solutions to identify relevant arousal patterns with special emphasis on robustness and artifacts tolerance. It is a multistage method that after obtaining an initial set of events, improves the detection finding common EEG arousal patterns. Finally, false positives are discarded after examining each candidate within the context of clinical definitions.

RESULTS

22 polysomnographic recordings from real patients were used to validate the method. The results obtained were encouraging, achieving a precision value of 0.86 and a F₁ score value of 0.79. When compared with the gold standard, the method achieves a substantial agreement (Kappa coefficient of 0.78), with an almost perfect agreement with ten recordings.

CONCLUSIONS

The algorithm designed achieved encouraging results and shows robust behavior in presence of signal artifacts. Its low-coupled design allows its implementation on different development platforms, and an easy combination with other methods.