Metric learning for automatic sleep stage classification

Phan, Huy; Quan, Do; Do, The-Luan; Vu, Duc-Lung

doi:10.1109/embc.2013.6610677

Cited by 30 publications

(6 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In shallow learning, meaningful features are extracted by humans drawing upon their knowledge of both signal processing and sleep stage scoring manual. The extracted features span in time, frequency, and time–frequency domains, including non-linear features 21 – 27 , 29 , 35 , 39 , 41 . Various classification models are utilized to classify these extracted features ranging from traditional machine learning models such as k-nearest neighbor (kNN) 35 , support vector machines (SVM) 21 , 27 , 33 , decision tree 25 , and random forest 24 , 29 , to artificial neural networks (ANNs), and deep learning models which excel in the feature-based approach 23 .…”

Section: Introductionmentioning

confidence: 99%

ZleepAnlystNet: a novel deep learning model for automatic sleep stage scoring based on single-channel raw EEG data using separating training

Jirakittayakorn,

Wongsawat,

Mitrirattanakul

2024

Sci Rep

View full text Add to dashboard Cite

Numerous models for sleep stage scoring utilizing single-channel raw EEG signal have typically employed CNN and BiLSTM architectures. While these models, incorporating temporal information for sequence classification, demonstrate superior overall performance, they often exhibit low per-class performance for N1-stage, necessitating an adjustment of loss function. However, the efficacy of such adjustment is constrained by the training process. In this study, a pioneering training approach called separating training is introduced, alongside a novel model, to enhance performance. The developed model comprises 15 CNN models with varying loss function weights for feature extraction and 1 BiLSTM for sequence classification. Due to its architecture, this model cannot be trained using an end-to-end approach, necessitating separate training for each component using the Sleep-EDF dataset. Achieving an overall accuracy of 87.02%, MF1 of 82.09%, Kappa of 0.8221, and per-class F1-socres (W 90.34%, N1 54.23%, N2 89.53%, N3 88.96%, and REM 87.40%), our model demonstrates promising performance. Comparison with sleep technicians reveals a Kappa of 0.7015, indicating alignment with reference sleep stags. Additionally, cross-dataset validation and adaptation through training with the SHHS dataset yield an overall accuracy of 84.40%, MF1 of 74.96% and Kappa of 0.7785 when tested with the Sleep-EDF-13 dataset. These findings underscore the generalization potential in model architecture design facilitated by our novel training approach.

show abstract

Section: Introductionmentioning

confidence: 99%

ZleepAnlystNet: a novel deep learning model for automatic sleep stage scoring based on single-channel raw EEG data using separating training

Jirakittayakorn,

Wongsawat,

Mitrirattanakul

2024

Sci Rep

View full text Add to dashboard Cite

show abstract

“…Machine learning typically involves feature extraction and classification. Phan et al [10] extracted features from single-channel EEG signals and used the k nearest neighbor algorithm with the Mahalanobis distance for classification. The method performs well in addressing four types of sleep stage problems.…”

Section: Introductionmentioning

confidence: 99%

Automatic Sleep Stage Classification Using a Taguchi-Based Multiscale Convolutional Compensatory Fuzzy Neural Network

Lin,

Lin

2023

Applied Sciences

View full text Add to dashboard Cite

Current methods for sleep stage detection rely on sensors to collect physiological data. These methods are inaccurate and take up considerable medical resources. Thus, in this study, we propose a Taguchi-based multiscale convolutional compensatory fuzzy neural network (T-MCCFNN) model to automatically detect and classify sleep stages. In the proposed T-MCCFNN model, multiscale convolution kernels extract features of the input electroencephalogram signal and a compensatory fuzzy neural network is used in place of a traditional fully connected network as a classifier to improve the convergence rate during learning and to reduce the number of model parameters required. Due to the complexity of general deep learning networks, trial and error methods are often used to determine their parameters. However, this method is very time-consuming. Therefore, this study uses the Taguchi method instead, where the optimal parameter combination is identified over a minimal number of experiments. We use the Sleep-EDF database to evaluate the proposed model. The results indicate that the proposed T-MCCFNN sleep stage classification accuracy is 85.3%, which is superior to methods proposed by other scholars.

show abstract

“…Extracted features have been classified using various algorithms, i.e. k-nearest neighbors [26,27], support vector machines [25,[28][29][30], ensemble classifiers [31][32][33], decision trees [27,34], artificial neural networks [25,29,35], and threshold-based classifiers [22][23][24]36]. These earlier works have demonstrated that automatic detection of arousals and sleep stages can be achieved by employing methods relying on handcrafted features.…”

Section: Introductionmentioning

confidence: 99%

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

Zan,

Yildiz

2023

J. Neural Eng.

View full text Add to dashboard Cite

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.

show abstract

Metric learning for automatic sleep stage classification

Cited by 30 publications

References 18 publications

ZleepAnlystNet: a novel deep learning model for automatic sleep stage scoring based on single-channel raw EEG data using separating training

ZleepAnlystNet: a novel deep learning model for automatic sleep stage scoring based on single-channel raw EEG data using separating training

Automatic Sleep Stage Classification Using a Taguchi-Based Multiscale Convolutional Compensatory Fuzzy Neural Network

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

Contact Info

Product

Resources

About