Sleep staging from electrocardiography and respiration with deep learning.
Ontology highlight
ABSTRACT:
SUBMITTER: Sun H
PROVIDER: S-EPMC7355395 | biostudies-literature | 2020 Jul
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Clinical sleep evaluations currently require multimodal data collection and manual review by human experts, making them expensive and unsuitable for longer term studies. Sleep staging using cardiac rhythm is an active area of research because it can be measured much more easily using a wide variety of both medical and consumer-grade devices. In this study, we applied deep learning methods to create an algorithm for automated sleep stage scoring using the instantaneous heart rate (IHR) time series extracted from the electrocardiogram (ECG). We trained and validated an algorithm on over 10,000 nights of data from the Sleep Heart Health Study (SHHS) and Multi-Ethnic Study of Atherosclerosis (MESA). The algorithm has an overall performance of 0.77 accuracy and 0.66 kappa against the reference stages on a held-out portion of the SHHS dataset for classifying every 30?s of sleep into four classes: wake, light sleep, deep sleep, and rapid eye movement (REM). Moreover, we demonstrate that the algorithm generalizes well to an independent dataset of 993 subjects labeled by American Academy of Sleep Medicine (AASM) licensed clinical staff at Massachusetts General Hospital that was not used for training or validation. Finally, we demonstrate that the stages predicted by our algorithm can reproduce previous clinical studies correlating sleep stages with comorbidities such as sleep apnea and hypertension as well as demographics such as age and gender.
| S-EPMC7441407 | biostudies-literature
Project description:Accurate identification of sleep stages is essential in the diagnosis of sleep disorders (e.g. obstructive sleep apnea [OSA]) but relies on labor-intensive electroencephalogram (EEG)-based manual scoring. Furthermore, long-term assessment of sleep relies on actigraphy differentiating only between wake and sleep periods without identifying specific sleep stages and having low reliability in identifying wake periods after sleep onset. To address these issues, we aimed to develop an automatic method for identifying the sleep stages from the photoplethysmogram (PPG) signal obtained with a simple finger pulse oximeter. PPG signals from the diagnostic polysomnographies of susptected OSA patients (n = 894) were utilized to develop a combined convolutional and recurrent neural network. The deep learning model was trained individually for three-stage (wake/NREM/REM), four-stage (wake/N1+N2/N3/REM), and five-stage (wake/N1/N2/N3/REM) classification of sleep. The three-stage model achieved an epoch-by-epoch accuracy of 80.1% with Cohen's κ of 0.65. The four- and five-stage models achieved 68.5% (κ = 0.54), and 64.1% (κ = 0.51) accuracies, respectively. With the five-stage model, the total sleep time was underestimated with a mean bias error (SD) of of 7.5 (55.2) minutes. The PPG-based deep learning model enabled accurate estimation of sleep time and differentiation between sleep stages with a moderate agreement to manual EEG-based scoring. As PPG is already included in ambulatory polygraphic recordings, applying the PPG-based sleep staging could improve their diagnostic value by enabling simple, low-cost, and reliable monitoring of sleep and help assess otherwise overlooked conditions such as REM-related OSA.
| S-EPMC7658638 | biostudies-literature
Project description:Develop a high-performing, automated sleep scoring algorithm that can be applied to long-term scalp electroencephalography (EEG) recordings. Using a clinical dataset of polysomnograms from 6,431 patients (MGH-PSG dataset), we trained a deep neural network to classify sleep stages based on scalp EEG data. The algorithm consists of a convolutional neural network for feature extraction, followed by a recurrent neural network that extracts temporal dependencies of sleep stages. The algorithm's inputs are four scalp EEG bipolar channels (F3-C3, C3-O1, F4-C4, and C4-O2), which can be derived from any standard PSG or scalp EEG recording. We initially trained the algorithm on the MGH-PSG dataset and used transfer learning to fine-tune it on a dataset of long-term (24-72 h) scalp EEG recordings from 112 patients (scalpEEG dataset). The algorithm achieved a Cohen's kappa of 0.74 on the MGH-PSG holdout testing set and cross-validated Cohen's kappa of 0.78 after optimization on the scalpEEG dataset. The algorithm also performed well on two publicly available PSG datasets, demonstrating high generalizability. Performance on all datasets was comparable to the inter-rater agreement of human sleep staging experts (Cohen's kappa ~ 0.75 ± 0.11). The algorithm's performance on long-term scalp EEGs was robust over a wide age range and across common EEG background abnormalities. We developed a deep learning algorithm that achieves human expert level sleep staging performance on long-term scalp EEG recordings. This algorithm, which we have made publicly available, greatly facilitates the use of large long-term EEG clinical datasets for sleep-related research.
| S-EPMC7686563 | biostudies-literature
Project description:Introduction: Automated sleep staging using deep learning models typically requires training on hundreds of sleep recordings, and pre-training on public databases is therefore common practice. However, suboptimal sleep stage performance may occur from mismatches between source and target datasets, such as differences in population characteristics (e.g., an unrepresented sleep disorder) or sensors (e.g., alternative channel locations for wearable EEG). Methods: We investigated three strategies for training an automated single-channel EEG sleep stager: pre-training (i.e., training on the original source dataset), training-from-scratch (i.e., training on the new target dataset), and fine-tuning (i.e., training on the original source dataset, fine-tuning on the new target dataset). As source dataset, we used the F3-M2 channel of healthy subjects (N = 94). Performance of the different training strategies was evaluated using Cohen's Kappa (κ) in eight smaller target datasets consisting of healthy subjects (N = 60), patients with obstructive sleep apnea (OSA, N = 60), insomnia (N = 60), and REM sleep behavioral disorder (RBD, N = 22), combined with two EEG channels, F3-M2 and F3-F4. Results: No differences in performance between the training strategies was observed in the age-matched F3-M2 datasets, with an average performance across strategies of κ = .83 in healthy, κ = .77 in insomnia, and κ = .74 in OSA subjects. However, in the RBD set, where data availability was limited, fine-tuning was the preferred method (κ = .67), with an average increase in κ of .15 to pre-training and training-from-scratch. In the presence of channel mismatches, targeted training is required, either through training-from-scratch or fine-tuning, increasing performance with κ = .17 on average. Discussion: We found that, when channel and/or population mismatches cause suboptimal sleep staging performance, a fine-tuning approach can yield similar to superior performance compared to building a model from scratch, while requiring a smaller sample size. In contrast to insomnia and OSA, RBD data contains characteristics, either inherent to the pathology or age-related, which apparently demand targeted training.
| S-EPMC10796543 | biostudies-literature
Project description:Background Severe, symptomatic aortic stenosis (AS) is associated with poor prognoses. However, early detection of AS is difficult because of the long asymptomatic period experienced by many patients, during which screening tools are ineffective. The aim of this study was to develop and validate a deep learning-based algorithm, combining a multilayer perceptron and convolutional neural network, for detecting significant AS using ECGs. Methods and Results This retrospective cohort study included adult patients who had undergone both ECG and echocardiography. A deep learning-based algorithm was developed using 39 371 ECGs. Internal validation of the algorithm was performed with 6453 ECGs from one hospital, and external validation was performed with 10 865 ECGs from another hospital. The end point was significant AS (beyond moderate). We used demographic information, features, and 500-Hz, 12-lead ECG raw data as predictive variables. In addition, we identified which region had the most significant effect on the decision-making of the algorithm using a sensitivity map. During internal and external validation, the areas under the receiver operating characteristic curve of the deep learning-based algorithm using 12-lead ECG for detecting significant AS were 0.884 (95% CI, 0.880-0.887) and 0.861 (95% CI, 0.858-0.863), respectively; those using a single-lead ECG signal were 0.845 (95% CI, 0.841-0.848) and 0.821 (95% CI, 0.816-0.825), respectively. The sensitivity map showed the algorithm focused on the T wave of the precordial lead to determine the presence of significant AS. Conclusions The deep learning-based algorithm demonstrated high accuracy for significant AS detection using both 12-lead and single-lead ECGs.
| S-EPMC7428650 | biostudies-literature
Project description:Background and objectivesScreening and early diagnosis for heart failure (HF) are critical. However, conventional screening diagnostic methods have limitations, and electrocardiography (ECG)-based HF identification may be helpful. This study aimed to develop and validate a deep-learning algorithm for ECG-based HF identification (DEHF).MethodsThe study involved 2 hospitals and 55,163 ECGs of 22,765 patients who performed echocardiography within 4 weeks were study subjects. ECGs were divided into derivation and validation data. Demographic and ECG features were used as predictive variables. The primary endpoint was detection of HF with reduced ejection fraction (HFrEF; ejection fraction [EF]?40%), and the secondary endpoint was HF with mid-range to reduced EF (?50%). We developed the DEHF using derivation data and the algorithm representing the risk of HF between 0 and 1. We confirmed accuracy and compared logistic regression (LR) and random forest (RF) analyses using validation data.ResultsThe area under the receiver operating characteristic curves (AUROCs) of DEHF for identification of HFrEF were 0.843 (95% confidence interval, 0.840-0.845) and 0.889 (0.887-0.891) for internal and external validation, respectively, and these results significantly outperformed those of LR (0.800 [0.797-0.803], 0.847 [0.844-0.850]) and RF (0.807 [0.804-0.810], 0.853 [0.850-0.855]) analyses. The AUROCs of deep learning for identification of the secondary endpoint was 0.821 (0.819-0.823) and 0.850 (0.848-0.852) for internal and external validation, respectively, and these results significantly outperformed those of LR and RF.ConclusionsThe deep-learning algorithm accurately identified HF using ECG features and outperformed other machine-learning methods.
| S-EPMC6597456 | biostudies-literature
Project description:BackgroundSerial electrocardiography aims to contribute to electrocardiogram (ECG) diagnosis by comparing the ECG under consideration with a previously made ECG in the same individual. Here, we present a novel algorithm to construct dedicated deep-learning neural networks (NNs) that are specialized in detecting newly emerging or aggravating existing cardiac pathology in serial ECGs.MethodsWe developed a novel deep-learning method for serial ECG analysis and tested its performance in detection of heart failure in post-infarction patients, and in the detection of ischemia in patients who underwent elective percutaneous coronary intervention. Core of the method is the repeated structuring and learning procedure that, when fed with 13 serial ECG difference features (intra-individual differences in: QRS duration; QT interval; QRS maximum; T-wave maximum; QRS integral; T-wave integral; QRS complexity; T-wave complexity; ventricular gradient; QRS-T spatial angle; heart rate; J-point amplitude; and T-wave symmetry), dynamically creates a NN of at most three hidden layers. An optimization process reduces the possibility of obtaining an inefficient NN due to adverse initialization.ResultsApplication of our method to the two clinical ECG databases yielded 3-layer NN architectures, both showing high testing performances (areas under the receiver operating curves were 84% and 83%, respectively).ConclusionsOur method was successful in two different clinical serial ECG applications. Further studies will investigate if other problem-specific NNs can successfully be constructed, and even if it will be possible to construct a universal NN to detect any pathologic ECG change.
| S-EPMC6371549 | biostudies-literature
Project description:Regular aerobic physical activity is of utmost importance in maintaining a good health status and preventing cardiovascular diseases (CVDs). Although cardiopulmonary exercise testing (CPX) is an essential examination for noninvasive estimation of ventilatory threshold (VT), defined as the clinically equivalent to aerobic exercise, its evaluation requires an expensive respiratory gas analyzer and expertize. To address these inconveniences, this study investigated the feasibility of a deep learning (DL) algorithm with single-lead electrocardiography (ECG) for estimating the aerobic exercise threshold. Two hundred sixty consecutive patients with CVDs who underwent CPX were analyzed. Single-lead ECG data were stored as time-series voltage data with a sampling rate of 1000 Hz. The data of preprocessed ECG and time point at VT calculated by respiratory gas analyzer were used to train a neural network. The trained model was applied on an independent test cohort, and the DL threshold (DLT; a time of VT estimated through the DL algorithm) was calculated. We compared the correlation between oxygen uptake of the VT (VT-VO2) and the DLT (DLT-VO2). Our DL model showed that the DLT-VO2 was confirmed to be significantly correlated with the VT-VO2 (r = 0.875; P < 0.001), and the mean difference was nonsignificant (-0.05 ml/kg/min, P > 0.05), which displayed strong agreements between the VT and the DLT. The DL algorithm using single-lead ECG data enabled accurate estimation of VT in patients with CVDs. The DL algorithm may be a novel way for estimating aerobic exercise threshold.
| S-EPMC7596490 | biostudies-literature
Project description:BACKGROUND:The detection of dyskalemias-hypokalemia and hyperkalemia-currently depends on laboratory tests. Since cardiac tissue is very sensitive to dyskalemia, electrocardiography (ECG) may be able to uncover clinically important dyskalemias before laboratory results. OBJECTIVE:Our study aimed to develop a deep-learning model, ECG12Net, to detect dyskalemias based on ECG presentations and to evaluate the logic and performance of this model. METHODS:Spanning from May 2011 to December 2016, 66,321 ECG records with corresponding serum potassium (K+) concentrations were obtained from 40,180 patients admitted to the emergency department. ECG12Net is an 82-layer convolutional neural network that estimates serum K+ concentration. Six clinicians-three emergency physicians and three cardiologists-participated in human-machine competition. Sensitivity, specificity, and balance accuracy were used to evaluate the performance of ECG12Net with that of these physicians. RESULTS:In a human-machine competition including 300 ECGs of different serum K+ concentrations, the area under the curve for detecting hypokalemia and hyperkalemia with ECG12Net was 0.926 and 0.958, respectively, which was significantly better than that of our best clinicians. Moreover, in detecting hypokalemia and hyperkalemia, the sensitivities were 96.7% and 83.3%, respectively, and the specificities were 93.3% and 97.8%, respectively. In a test set including 13,222 ECGs, ECG12Net had a similar performance in terms of sensitivity for severe hypokalemia (95.6%) and severe hyperkalemia (84.5%), with a mean absolute error of 0.531. The specificities for detecting hypokalemia and hyperkalemia were 81.6% and 96.0%, respectively. CONCLUSIONS:A deep-learning model based on a 12-lead ECG may help physicians promptly recognize severe dyskalemias and thereby potentially reduce cardiac events.
| S-EPMC7082733 | biostudies-literature
Project description:BackgroundOur aim was to establish a deep learning radiomics method to preoperatively evaluate regional lymph node (LN) staging for hilar cholangiocarcinoma (HC) patients.Methods and materialsOf the 179 enrolled HC patients, 90 were pathologically diagnosed with lymph node metastasis. Quantitative radiomic features and deep learning features were extracted. An LN metastasis status classifier was developed through integrating support vector machine, high-performance deep learning radiomics signature, and three clinical characteristics. An LN metastasis stratification classifier (N1 vs. N2) was also proposed with subgroup analysis.ResultsThe average areas under the receiver operating characteristic curve (AUCs) of the LN metastasis status classifier reached 0.866 in the training cohort and 0.870 in the external test cohorts. Meanwhile, the LN metastasis stratification classifier performed well in predicting the risk of LN metastasis, with an average AUC of 0.946.ConclusionsTwo classifiers derived from computed tomography images performed well in predicting LN staging in HC and will be reliable evaluation tools to improve decision-making.
| S-EPMC8576333 | biostudies-literature