Expert-level automated sleep staging of long-term scalp electroencephalography recordings using deep learning

Hyperarousal features in the sleep architecture of individuals with and without insomnia

Sleep architecture encodes relevant information on the structure of sleep and has been used to assess hyperarousal in insomnia. This study investigated whether polysomnography-derived sleep architecture displays signs of hyperarousal in individuals with insomnia compared with individuals without insomnia. Data from Phase 3 clinical trials, private clinics and a cohort study were analysed. A comprehensive set of sleep architecture features previously associated with hyperarousal were retrospectively analysed focusing on sleep-wake transition probabilities, electroencephalographic spectra and sleep spindles, and enriched with a novel machine learning algorithm called the Wake Electroencephalographic Similarity Index. This analysis included 1710 individuals with insomnia and 1455 individuals without insomnia. Results indicate that individuals with insomnia had a higher likelihood of waking from all sleep stages, and showed increased relative alpha during Wake and N1 sleep and increased theta power during Wake when compared with individuals without insomnia. Relative delta power was decreased and Wake Electroencephalographic Similarity Index scores were elevated across all sleep stages except N3, suggesting more wake-like activity during these stages in individuals with insomnia. Additionally, sleep spindle density was decreased, and spindle dispersion was increased in individuals with insomnia. These findings suggest that insomnia is characterized by a dysfunction in sleep quality with a continuous hyperarousal, evidenced by changes in sleep-wake architecture.

Sleep functional connectivity, hyperexcitability, and cognition in Alzheimer's disease

INTRODUCTION

Sleep disturbances are common in Alzheimer's disease (AD) and may reflect pathologic changes in brain networks. To date, no studies have examined changes in sleep functional connectivity (FC) in AD or their relationship with network hyperexcitability and cognition.

METHODS

We assessed electroencephalogram (EEG) sleep FC in 33 healthy controls, 36 individuals with AD without epilepsy, and 14 individuals with AD and epilepsy.

RESULTS

AD participants showed increased gamma connectivity in stage 2 sleep (N2), which was associated with longitudinal cognitive decline. Network hyperexcitability in AD was associated with a distinct sleep connectivity signature, characterized by decreased N2 delta connectivity and reversal of several connectivity changes associated with AD. Machine learning algorithms using sleep connectivity features accurately distinguished diagnostic groups and identified "fast cognitive decliners" among study participants who had AD.

DISCUSSION

Our findings reveal changes in sleep functional networks associated with cognitive decline in AD and may have implications for disease monitoring and therapeutic development.

HIGHLIGHTS

Brain functional connectivity (FC) in Alzheimer's disease is altered during sleep. Sleep FC measures correlate with cognitive decline in AD. Network hyperexcitability in AD has a distinct sleep connectivity signature.

Explainable AI-driven model for gastrointestinal cancer classification

Although the detection procedure has been shown to be highly effective, there are several obstacles to overcome in the usage of AI-assisted cancer cell detection in clinical settings. These issues stem mostly from the failure to identify the underlying processes. Because AI-assisted diagnosis does not offer a clear decision-making process, doctors are dubious about it. In this instance, the advent of Explainable Artificial Intelligence (XAI), which offers explanations for prediction models, solves the AI black box issue. The SHapley Additive exPlanations (SHAP) approach, which results in the interpretation of model predictions, is the main emphasis of this work. The intermediate layer in this study was a hybrid model made up of three Convolutional Neural Networks (CNNs) (InceptionV3, InceptionResNetV2, and VGG16) that combined their predictions. The KvasirV2 dataset, which comprises pathological symptoms associated to cancer, was used to train the model. Our combined model yielded an accuracy of 93.17% and an F1 score of 97%. After training the combined model, we use SHAP to analyze images from these three groups to provide an explanation of the decision that affects the model prediction.

Research and application of deep learning-based sleep staging: Data, modeling, validation, and clinical practice

Over the past few decades, researchers have attempted to simplify and accelerate the process of sleep stage classification through various approaches; however, only a few such approaches have gained widespread acceptance. Artificial intelligence technology, particularly deep learning, is promising for earning the trust of the sleep medicine community in automated sleep-staging systems, thus facilitating its application in clinical practice and integration into daily life. We aimed to comprehensively review the latest methods that are applying deep learning for enhancing sleep staging efficiency and accuracy. Starting from the requisite "data" for constructing deep learning algorithms, we elucidated the current landscape of this domain and summarized the fundamental modeling process, encompassing signal selection, data pre-processing, model architecture, classification tasks, and performance metrics. Furthermore, we reviewed the applications of automated sleep staging in scenarios such as sleep-disorder screening, diagnostic procedures, and health monitoring and management. Finally, we conducted an in-depth analysis and discussion of the challenges and future in intelligent sleep staging, particularly focusing on large-scale sleep datasets, interdisciplinary collaborations, and human-computer interactions.

Artificial Intelligence Models for the Automation of Standard Diagnostics in Sleep Medicine-A Systematic Review

Sleep disorders, prevalent in the general population, present significant health challenges. The current diagnostic approach, based on a manual analysis of overnight polysomnograms (PSGs), is costly and time-consuming. Artificial intelligence has emerged as a promising tool in this context, offering a more accessible and personalized approach to diagnosis, particularly beneficial for under-served populations. This is a systematic review of AI-based models for sleep disorder diagnostics that were trained, validated, and tested on diverse clinical datasets. An extensive search of PubMed and IEEE databases yielded 2114 articles, but only 18 met our stringent selection criteria, underscoring the scarcity of thoroughly validated AI models in sleep medicine. The findings emphasize the necessity of a rigorous validation of AI models on multimodal clinical data, a step crucial for their integration into clinical practice. This would be in line with the American Academy of Sleep Medicine's support of AI research.

Sustained polyphasic sleep restriction abolishes human growth hormone release

STUDY OBJECTIVES

Voluntary sleep restriction is a common phenomenon in industrialized societies aiming to increase time spent awake and thus productivity. We explored how restricting sleep to a radically polyphasic schedule affects neural, cognitive, and endocrine characteristics.

METHODS

Ten young healthy participants were restricted to one 20-minute nap opportunity at the end of every 4 hours (i.e. six sleep episodes per 24 hours) without any extended core sleep window, which resulted in a cumulative sleep amount of just 2 hours per day (i.e. ~20 minutes per bout).

RESULTS

All but one participant terminated this schedule during the first month. The remaining participant (a 25-year-old male) succeeded in adhering to a polyphasic schedule for five out of the eight planned weeks. Cognitive and psychiatric measures showed modest changes during polyphasic as compared to monophasic sleep, while in-blood cortisol or melatonin release patterns and amounts were apparently unaltered. In contrast, growth hormone release was almost entirely abolished (>95% decrease), with the residual release showing a considerably changed polyphasic secretional pattern.

CONCLUSIONS

Even though the study was initiated by volunteers with exceptional intrinsic motivation and commitment, none of them could tolerate the intended 8 weeks of the polyphasic schedule. Considering the decreased vigilance, abolished growth hormone release, and neurophysiological sleep changes observed, it is doubtful that radically polyphasic sleep schedules can subserve the different functions of sleep to a sufficient degree.

Survey of Transfer Learning Approaches in the Machine Learning of Digital Health Sensing Data

Machine learning and digital health sensing data have led to numerous research achievements aimed at improving digital health technology. However, using machine learning in digital health poses challenges related to data availability, such as incomplete, unstructured, and fragmented data, as well as issues related to data privacy, security, and data format standardization. Furthermore, there is a risk of bias and discrimination in machine learning models. Thus, developing an accurate prediction model from scratch can be an expensive and complicated task that often requires extensive experiments and complex computations. Transfer learning methods have emerged as a feasible solution to address these issues by transferring knowledge from a previously trained task to develop high-performance prediction models for a new task. This survey paper provides a comprehensive study of the effectiveness of transfer learning for digital health applications to enhance the accuracy and efficiency of diagnoses and prognoses, as well as to improve healthcare services. The first part of this survey paper presents and discusses the most common digital health sensing technologies as valuable data resources for machine learning applications, including transfer learning. The second part discusses the meaning of transfer learning, clarifying the categories and types of knowledge transfer. It also explains transfer learning methods and strategies, and their role in addressing the challenges in developing accurate machine learning models, specifically on digital health sensing data. These methods include feature extraction, fine-tuning, domain adaptation, multitask learning, federated learning, and few-/single-/zero-shot learning. This survey paper highlights the key features of each transfer learning method and strategy, and discusses the limitations and challenges of using transfer learning for digital health applications. Overall, this paper is a comprehensive survey of transfer learning methods on digital health sensing data which aims to inspire researchers to gain knowledge of transfer learning approaches and their applications in digital health, enhance the current transfer learning approaches in digital health, develop new transfer learning strategies to overcome the current limitations, and apply them to a variety of digital health technologies.

Standardized image-based polysomnography database and deep learning algorithm for sleep-stage classification

STUDY OBJECTIVES

Polysomnography (PSG) scoring is labor-intensive, subjective, and often ambiguous. Recently several deep learning (DL) models for automated sleep scoring have been developed, they are tied to a fixed amount of input channels and resolution. In this study, we constructed a standardized image-based PSG dataset in order to overcome the heterogeneity of raw signal data obtained from various PSG devices and various sleep laboratory environments.

METHODS

All individually exported European data format files containing raw signals were converted into images with an annotation file, which contained the demographics, diagnoses, and sleep statistics. An image-based DL model for automatic sleep staging was developed, compared with a signal-based model, and validated in an external dataset.

RESULTS

We constructed 10253 image-based PSG datasets using a standardized format. Among these, 7745 diagnostic PSG data were used to develop our DL model. The DL model using the image dataset showed similar performance to the signal-based dataset for the same subject. The overall DL accuracy was greater than 80%, even with severe obstructive sleep apnea. Moreover, for the first time, we showed explainable DL in the field of sleep medicine as visualized key inference regions using Eigen-class activation maps. Furthermore, when a DL model for sleep scoring performs external validation, we achieved a relatively good performance.

CONCLUSIONS

Our main contribution demonstrates the availability of a standardized image-based dataset, and highlights that changing the data sampling rate or number of sensors may not require retraining, although performance decreases slightly as the number of sensors decreases.

Altered Sleep Microarchitecture and Cognitive Impairment in Patients With Temporal Lobe Epilepsy

BACKGROUND AND OBJECTIVES

To investigate the spatiotemporal characteristics of sleep waveforms in temporal lobe epilepsy (TLE) and examine their association with cognition.

METHODS

In this retrospective, cross-sectional study, we examined overnight EEG data from adult patients with TLE and nonepilepsy comparisons (NECs) admitted to the epilepsy monitoring unit at Mass General Brigham hospitals. Automated algorithms were used to characterize sleep macroarchitecture (sleep stages) and microarchitecture (spindles, slow oscillations [SOs]) on scalp EEG and to detect hippocampal interictal epileptiform discharges (hIEDs) from foramen ovale electrodes simultaneously recorded in a subset of patients with TLE. We examined the association of sleep features and hIEDs with memory and executive function from clinical neuropsychological evaluations.

RESULTS

A total of 81 adult patients with TLE and 28 NEC adult patients were included with similar mean ages. There were no significant differences in sleep macroarchitecture between groups, including relative time spent in each sleep stage, sleep efficiency, and sleep fragmentation. By contrast, the spatiotemporal characteristics of sleep microarchitecture were altered in TLE compared with NEC and were associated with cognitive impairments. Specifically, we observed a ∼30% reduction in spindle density in patients with TLE compared with NEC, which was significantly associated with worse memory performance. Spindle-SO coupling strength was also reduced in TLE and, in contrast to spindles, was associated with diminished executive function. We found no significant association between sleep macroarchitectural and microarchitectural parameters and hIEDs.

DISCUSSION

There is a fundamental alteration of sleep microarchitecture in TLE, characterized by a reduction in spindle density and spindle-SO coupling, and these changes may contribute to neurocognitive comorbidity in this disorder.

Challenges of Applying Automated Polysomnography Scoring at Scale

Automatic polysomnography analysis can be leveraged to shorten scoring times, reduce associated costs, and ultimately improve the overall diagnosis of sleep disorders. Multiple and diverse strategies have been attempted for implementation of this technology at scale in the routine workflow of sleep centers. The field, however, is complex and presents unsolved challenges in a number of areas. Recent developments in computer science and artificial intelligence are nevertheless closing the gap. Technological advances are also opening new pathways for expanding our current understanding of the domain and its analysis.

Current status and prospects of automatic sleep stages scoring: Review

The scoring of sleep stages is one of the essential tasks in sleep analysis. Since a manual procedure requires considerable human and financial resources, and incorporates some subjectivity, an automated approach could result in several advantages. There have been many developments in this area, and in order to provide a comprehensive overview, it is essential to review relevant recent works and summarise the characteristics of the approaches, which is the main aim of this article. To achieve it, we examined articles published between 2018 and 2022 that dealt with the automated scoring of sleep stages. In the final selection for in-depth analysis, 125 articles were included after reviewing a total of 515 publications. The results revealed that automatic scoring demonstrates good quality (with Cohen's kappa up to over 0.80 and accuracy up to over 90%) in analysing EEG/EEG + EOG + EMG signals. At the same time, it should be noted that there has been no breakthrough in the quality of results using these signals in recent years. Systems involving other signals that could potentially be acquired more conveniently for the user (e.g. respiratory, cardiac or movement signals) remain more challenging in the implementation with a high level of reliability but have considerable innovation capability. In general, automatic sleep stage scoring has excellent potential to assist medical professionals while providing an objective assessment.

Automatic sleep staging of single-channel EEG based on domain adversarial neural networks and domain self-attention

The diagnosis and management of sleep problems depend heavily on sleep staging. For autonomous sleep staging, many data-driven deep learning models have been presented by trying to construct a large-labeled auxiliary sleep dataset and test it by electroencephalograms on different subjects. These approaches suffer a significant setback cause it assumes the training and test data come from the same or similar distribution. However, this is almost impossible in scenario cross-dataset due to inherent domain shift between domains. Unsupervised domain adaption was recently created to address the domain shift issue. However, only a few customized UDA solutions for sleep staging due to two limitations in previous UDA methods. First, the domain classifier does not consider boundaries between classes. Second, they depend on a shared model to align the domain that could miss the information of domains when extracting features. Given those restrictions, we present a novel UDA approach that combines category decision boundaries and domain discriminator to align the distributions of source and target domains. Also, to keep the domain-specific features, we create an unshared attention method. In addition, we investigated effective data augmentation in cross-dataset sleep scenarios. The experimental results on three datasets validate the efficacy of our approach and show that the proposed method is superior to state-of-the-art UDA methods on accuracy and MF1-Score.

Rating by detection: an artifact detection protocol for rating EEG quality with average event duration

Objective. Quantitative evaluation protocols are critical for the development of algorithms that remove artifacts from real electroencephalography (EEG) optimally. However, visually inspecting the real EEG to select the top-performing artifact removal pipeline is infeasible while hand-crafted EEG data allow assessing artifact removal configurations only in a simulated environment. This study proposes a novel, principled approach for quantitatively evaluating algorithmically corrected EEG without access to ground truth in real-world conditions.Approach. Our offline evaluation protocol uses a detector to score the presence of artifacts. It computes their average duration, which measures the recovered EEG's deviation from the modeled background activity with a single score. As we expect the detector to make generalization errors, we employ a generic and configurable Wiener-based artifact removal method to validate the reliability of our detection protocol.Main results. Quantitative experiments extensively compare many Wiener filters and show their consistent rankings agree with their theoretical assumptions and expectations.Significance. The rating-by-detection protocol with the average event duration measure should be of value for EEG practitioners and developers. After removing artifacts from real EEG, the protocol experimentally shows that reliable comparisons between many artifact filtering configurations are possible despite the missing ground-truth neural signals.

Clinical applications of artificial intelligence in sleep medicine: a sleep clinician's perspective

BACKGROUND

The past few years have seen a rapid emergence of artificial intelligence (AI)-enabled technology in the field of sleep medicine. AI refers to the capability of computer systems to perform tasks conventionally considered to require human intelligence, such as speech recognition, decision-making, and visual recognition of patterns and objects. The practice of sleep tracking and measuring physiological signals in sleep is widely practiced. Therefore, sleep monitoring in both the laboratory and ambulatory environments results in the accrual of massive amounts of data that uniquely positions the field of sleep medicine to gain from AI.

METHOD

The purpose of this article is to provide a concise overview of relevant terminology, definitions, and use cases of AI in sleep medicine. This was supplemented by a thorough review of relevant published literature.

RESULTS

Artificial intelligence has several applications in sleep medicine including sleep and respiratory event scoring in the sleep laboratory, diagnosing and managing sleep disorders, and population health. While still in its nascent stage, there are several challenges which preclude AI's generalizability and wide-reaching clinical applications. Overcoming these challenges will help integrate AI seamlessly within sleep medicine and augment clinical practice.

CONCLUSION

Artificial intelligence is a powerful tool in healthcare that may improve patient care, enhance diagnostic abilities, and augment the management of sleep disorders. However, there is a need to regulate and standardize existing machine learning algorithms prior to its inclusion in the sleep clinic.

Scoring sleep with artificial intelligence enables quantification of sleep stage ambiguity: hypnodensity based on multiple expert scorers and auto-scoring

STUDY OBJECTIVES

To quantify the amount of sleep stage ambiguity across expert scorers and to validate a new auto-scoring platform against sleep staging performed by multiple scorers.

METHODS

We applied a new auto-scoring system to three datasets containing 95 PSGs scored by 6-12 scorers, to compare sleep stage probabilities (hypnodensity; i.e. the probability of each sleep stage being assigned to a given epoch) as the primary output, as well as a single sleep stage per epoch assigned by hierarchical majority rule.

RESULTS

The percentage of epochs with 100% agreement across scorers was 46 ± 9%, 38 ± 10% and 32 ± 9% for the datasets with 6, 9, and 12 scorers, respectively. The mean intra-class correlation coefficient between sleep stage probabilities from auto- and manual-scoring was 0.91, representing excellent reliability. Within each dataset, agreement between auto-scoring and consensus manual-scoring was significantly higher than agreement between manual-scoring and consensus manual-scoring (0.78 vs. 0.69; 0.74 vs. 0.67; and 0.75 vs. 0.67; all p < 0.01).

CONCLUSIONS

Analysis of scoring performed by multiple scorers reveals that sleep stage ambiguity is the rule rather than the exception. Probabilities of the sleep stages determined by artificial intelligence auto-scoring provide an excellent estimate of this ambiguity. Compared to consensus manual-scoring, sleep staging derived from auto-scoring is for each individual PSG noninferior to manual-scoring meaning that auto-scoring output is ready for interpretation without the need for manual adjustment.

EEG Entropy in REM Sleep as a Physiologic Biomarker in Early Clinical Stages of Alzheimer's Disease

BACKGROUND

Alzheimer's disease (AD) is associated with EEG changes across the sleep-wake cycle. As the brain is a non-linear system, non-linear EEG features across behavioral states may provide an informative physiologic biomarker of AD. Multiscale fluctuation dispersion entropy (MFDE) provides a sensitive non-linear measure of EEG information content across a range of biologically relevant time-scales.

OBJECTIVE

To evaluate MFDE in awake and sleep EEGs as a potential biomarker for AD.

METHODS

We analyzed overnight scalp EEGs from 35 cognitively normal healthy controls, 23 participants with mild cognitive impairment (MCI), and 19 participants with mild dementia due to AD. We examined measures of entropy in wake and sleep states, including a slow-to-fast-activity ratio of entropy (SFAR-entropy). We compared SFAR-entropy to linear EEG measures including a slow-to-fast-activity ratio of power spectral density (SFAR-PSD) and relative alpha power, as well as to cognitive function.

RESULTS

SFAR-entropy differentiated dementia from MCI and controls. This effect was greatest in REM sleep, a state associated with high cholinergic activity. Differentiation was evident in the whole brain EEG and was most prominent in temporal and occipital regions. Five minutes of REM sleep was sufficient to distinguish dementia from MCI and controls. Higher SFAR-entropy during REM sleep was associated with worse performance on the Montreal Cognitive Assessment. Classifiers based on REM sleep SFAR-entropy distinguished dementia from MCI and controls with high accuracy, and outperformed classifiers based on SFAR-PSD and relative alpha power.

CONCLUSION

SFAR-entropy measured in REM sleep robustly discriminates dementia in AD from MCI and healthy controls.

Acupuncture combined with cognitive-behavioural therapy for insomnia (CBT-I) in patients with insomnia: study protocol for a randomised controlled trial

INTRODUCTION

Insomnia affects physical and mental health due to the lack of continuous and complete sleep architecture. Polysomnograms (PSGs) are used to record electrical information to perform sleep architecture using deep learning. Although acupuncture combined with cognitive-behavioural therapy for insomnia (CBT-I) could not only improve sleep quality, solve anxiety, depression but also ameliorate poor sleep habits and detrimental cognition. Therefore, this study will focus on the effects of electroacupuncture combined with CBT-I on sleep architecture with deep learning.

METHODS AND ANALYSIS

This randomised controlled trial will evaluate the efficacy and effectiveness of electroacupuncture combined with CBT-I in patients with insomnia. Participants will be randomised to receive either electroacupuncture combined with CBT-I or sham acupuncture combined with CBT-I and followed up for 4 weeks. The primary outcome is sleep quality, which is evaluated by the Pittsburgh Sleep Quality Index. The secondary outcome measures include a measurement of depression severity, anxiety, maladaptive cognitions associated with sleep and adverse events. Sleep architecture will be assessed using deep learning on PSGs.

ETHICS AND DISSEMINATION

This trial has been approved by the institutional review boards and ethics committees of the First Affiliated Hospital of Sun Yat-sun University (2021763). The results will be disseminated through peer-reviewed journals. The results of this trial will be disseminated through peer-reviewed publications and conference abstracts or posters.

TRIAL REGISTRATION NUMBER

CTR2100052502.

Deep learning in neuroimaging data analysis: Applications, challenges, and solutions

Methods for the analysis of neuroimaging data have advanced significantly since the beginning of neuroscience as a scientific discipline. Today, sophisticated statistical procedures allow us to examine complex multivariate patterns, however most of them are still constrained by assuming inherent linearity of neural processes. Here, we discuss a group of machine learning methods, called deep learning, which have drawn much attention in and outside the field of neuroscience in recent years and hold the potential to surpass the mentioned limitations. Firstly, we describe and explain the essential concepts in deep learning: the structure and the computational operations that allow deep models to learn. After that, we move to the most common applications of deep learning in neuroimaging data analysis: prediction of outcome, interpretation of internal representations, generation of synthetic data and segmentation. In the next section we present issues that deep learning poses, which concerns multidimensionality and multimodality of data, overfitting and computational cost, and propose possible solutions. Lastly, we discuss the current reach of DL usage in all the common applications in neuroimaging data analysis, where we consider the promise of multimodality, capability of processing raw data, and advanced visualization strategies. We identify research gaps, such as focusing on a limited number of criterion variables and the lack of a well-defined strategy for choosing architecture and hyperparameters. Furthermore, we talk about the possibility of conducting research with constructs that have been ignored so far or/and moving toward frameworks, such as RDoC, the potential of transfer learning and generation of synthetic data.

A machine learning eye movement detection algorithm using electrooculography

Study Objectives

Eye movement quantification in polysomnograms (PSG) is difficult and resource intensive. Automated eye movement detection would enable further study of eye movement patterns in normal and abnormal sleep, which could be clinically diagnostic of neurologic disorders, or used to monitor potential treatments. We trained a long short-term memory (LSTM) algorithm that can identify eye movement occurrence with high sensitivity and specificity.

Methods

We conducted a retrospective, single-center study using one-hour PSG samples from 47 patients 18–90 years of age. Team members manually identified and trained an LSTM algorithm to detect eye movement presence, direction, and speed. We performed a 5-fold cross validation and implemented a “fuzzy” evaluation method to account for misclassification in the preceding and subsequent 1-second of gold standard manually labeled eye movements. We assessed G-means, discrimination, sensitivity, and specificity.

Results

Overall, eye movements occurred in 9.4% of the analyzed EOG recording time from 47 patients. Eye movements were present 3.2% of N2 (lighter stages of sleep) time, 2.9% of N3 (deep sleep), and 19.8% of REM sleep. Our LSTM model had average sensitivity of 0.88 and specificity of 0.89 in 5-fold cross validation, which improved to 0.93 and 0.92 respectively using the fuzzy evaluation scheme.

Conclusion

An automated algorithm can detect eye movements from EOG with excellent sensitivity and specificity. Noninvasive, automated eye movement detection has several potential clinical implications in improving sleep study stage classification and establishing normal eye movement distributions in healthy and unhealthy sleep, and in patients with and without brain injury.

An automated heart rate-based algorithm for sleep stage classification: Validation using conventional polysomnography and an innovative wearable electrocardiogram device

Background

The rapid advancement in wearable solutions to monitor and score sleep staging has enabled monitoring outside of the conventional clinical settings. However, most of the devices and algorithms lack extensive and independent validation, a fundamental step to ensure robustness, stability, and replicability of the results beyond the training and testing phases. These systems are thought not to be feasible and reliable alternatives to the gold standard, polysomnography (PSG).

Materials and methods

This validation study highlights the accuracy and precision of the proposed heart rate (HR)-based deep-learning algorithm for sleep staging. The illustrated solution can perform classification at 2-levels (Wake; Sleep), 3-levels (Wake; NREM; REM) or 4- levels (Wake; Light; Deep; REM) in 30-s epochs. The algorithm was validated using an open-source dataset of PSG recordings (Physionet CinC dataset, n = 994 participants, 994 recordings) and a proprietary dataset of ECG recordings (Z3Pulse, n = 52 participants, 112 recordings) collected with a chest-worn, wireless sensor and simultaneous PSG collection using SOMNOtouch.

Results

We evaluated the performance of the models in both datasets in terms of Accuracy (A), Cohen's kappa (K), Sensitivity (SE), Specificity (SP), Positive Predictive Value (PPV), and Negative Predicted Value (NPV). In the CinC dataset, the highest value of accuracy was achieved by the 2-levels model (0.8797), while the 3-levels model obtained the best value of K (0.6025). The 4-levels model obtained the lowest SE (0.3812) and the highest SP (0.9744) for the classification of Deep sleep segments. AHI and biological sex did not affect scoring, while a significant decrease of performance by age was reported across the models. In the Z3Pulse dataset, the highest value of accuracy was achieved by the 2-levels model (0.8812), whereas the 3-levels model obtained the best value of K (0.611). For classification of the sleep states, the lowest SE (0.6163) and the highest SP (0.9606) were obtained for the classification of Deep sleep segment.

Conclusion

The results of the validation procedure demonstrated the feasibility of accurate HR-based sleep staging. The combination of the proposed sleep staging algorithm with an inexpensive HR device, provides a cost-effective and non-invasive solution deployable in the home environment and robust across age, sex, and AHI scores.

A-PASS: An automated pipeline to analyze simultaneously acquired EEG-fMRI data for studying brain activities during sleep

OBJECTIVE

Concurrent electroencephalography and functional magnetic resonance imaging (EEG-fMRI) signals can be used to uncover the nature of brain activities during sleep. However, analyzing simultaneously acquired EEG-fMRI data is extremely time consuming and experience dependent. Thus, we developed a pipeline, which we named A-PASS, to automatically analyze simultaneously acquired EEG-fMRI data for studying brain activities during sleep.

APPROACH

A deep learning model was trained on a sleep EEG-fMRI dataset from 45 subjects and used to perform sleep stage scoring. Various fMRI indices can be calculated with A-PASS to depict the neurophysiological characteristics across different sleep stages. We tested the performance of A-PASS on an independent sleep EEG-fMRI dataset from 28 subjects. Statistical maps regarding the main effect of sleep stages and differences between each pair of stages of fMRI indices were generated and compared using both A-PASS and manual processing methods.

MAIN RESULTS

The deep learning model implemented in A-PASS achieved both an accuracy and F1-score higher than 70% for sleep stage classification on EEG data acquired during fMRI scanning. The statistical maps generated from A-PASS largely resembled those produced from manually scored stages plus a combination of multiple software programs.

SIGNIFICANCE

A-PASS allowed efficient EEG-fMRI data processing without manual operation and could serve as a reliable and powerful tool for simultaneous EEG-fMRI studies on sleep.

Simple and Autonomous Sleep Signal Processing System for the Detection of Obstructive Sleep Apneas

Obstructive sleep apnea (OSA) is a common sleep disorder characterized by repetitive upper airway obstruction, intermittent hypoxemia, and recurrent awakenings during sleep. The most used treatment for this syndrome is a device that generates a positive airway pressure—Continuous Positive Airway Pressure (CPAP), but it works continuously, whether or not there is apnea. An alternative consists on systems that detect apnea episodes and produce a stimulus that eliminates them. Article focuses on the development of a simple and autonomous processing system for the detection of obstructive sleep apneas, using polysomnography (PSG) signals: electroencephalography (EEG), electromyography (EMG), respiratory effort (RE), respiratory flow (RF), and oxygen saturation (SO2). The system is evaluated using, as a gold standard, 20 PSG tests labeled by sleep experts and it performs two analyses. A first analysis detects awake/sleep stages and is based on the accumulated amplitude in a channel-dependent frequency range, according to the criteria of the American Academy of Sleep Medicine (AASM). The second analysis detects hypopneas and apneas, based on analysis of the breathing cycle and oxygen saturation. The results show a good estimation of sleep events, where for 75% of the cases of patients analyzed it is possible to determine the awake/asleep states with an effectiveness of >92% and apneas and hypopneas with an effectiveness of >55%, through a simple processing system that could be implemented in an electronic device to be used in possible OSA treatments.

Noninvasive Detection of Hippocampal Epileptiform Activity on Scalp Electroencephalogram

Importance

The hippocampus is a highly epileptogenic brain region, yet over 90% of hippocampal epileptiform activity (HEA) cannot be identified on scalp electroencephalogram (EEG) by human experts. Currently, detection of HEA requires intracranial electrodes, which limits our understanding of the role of HEA in brain diseases.

Objective

To develop and validate a machine learning algorithm that accurately detects HEA from a standard scalp EEG, without the need for intracranial electrodes.

Design, Setting, and Participants

In this diagnostic study, conducted from 2008 to 2021, EEG data were used from patients with temporal lobe epilepsy (TLE) and healthy controls (HCs) to train and validate a deep neural network, HEAnet, to detect HEA on scalp EEG. Participants were evaluated at tertiary-level epilepsy centers at 2 academic hospitals: Massachusetts General Hospital (MGH) or Brigham and Women's Hospital (BWH). Included in the study were patients aged 12 to 78 years with a clinical diagnosis of TLE and HCs without epilepsy. Patients with TLE and HCs with a history of intracranial surgery were excluded from the study.

Exposures

Simultaneous intracranial EEG and/or scalp EEG.

Main Outcomes and Measures

Performance was assessed using cross-validated areas under the receiver operating characteristic curve (AUC ROC) and precision-recall curve (AUC PR) and additional clinically relevant metrics.

Results

HEAnet was trained and validated using data sets that were derived from a convenience sample of 141 eligible participants (97 with TLE and 44 HCs without epilepsy) whose retrospective EEG data were readily available. Data set 1 included the simultaneous scalp EEG and intracranial electrode recordings of 51 patients with TLE (mean [SD] age, 40.7 [15.9] years; 30 men [59%]) at MGH. An automatically generated training data set with 972 095 positive HEA examples was created, in addition to a held-out expert-annotated testing data set with 22 762 positive HEA examples. HEAnet's performance was validated on 2 independent scalp EEG data sets: (1) data set 2 (at MGH; 24 patients with TLE and 20 HCs; mean [SD] age, 42.3 [16.2] years; 17 men [39%]) and (2) data set 3 (at BWH; 22 patients with TLE and 24 HCs; mean [SD] age, 43.0 [14.4] years; 20 men [43%]). For single-event detection of HEA on data set 1, HEAnet achieved a mean (SD) AUC ROC of 0.89 (0.01) and a mean (SD) AUC PR of 0.39 (0.03). On external validation with data sets 2 and 3, HEAnet accurately distinguished TLE from HC (AUC ROC of 0.88 and 0.95, respectively) and predicted epilepsy lateralization with 100% and 92% accuracy, respectively. HEAnet tracked dynamic changes in HEA in response to seizure medication adjustments and performed comparably with human experts in diagnosing TLE from 1-hour scalp EEG recordings, diagnosing TLE in several individuals that experts missed. Without reducing specificity, addition of HEAnet to human expert EEG review increased sensitivity for diagnosing TLE in humans from 50% to 58% to 63% to 67%.

Conclusions and Relevance

Results of this diagnostic study suggest that HEAnet provides a novel, noninvasive, quantitative, and clinically relevant biomarker of hippocampal hyperexcitability in humans.

Polysomnographic identification of anxiety and depression using deep learning

Anxiety and depression are common psychiatric conditions associated with significant morbidity and healthcare costs. Sleep is an evolutionarily conserved health state. Anxiety and depression have a bidirectional relationship with sleep. This study reports on the use of analysis of polysomnographic data using deep learning methods to detect the presence of anxiety and depression. Polysomnography data on 940 patients performed at an academic sleep center during the 3-year period from 01/01/2016 to 12/31/2018 were identified for analysis. The data were divided into 3 subgroups: 205 patients with Anxiety/Depression, 349 patients with no Anxiety/Depression, and 386 patients with likely Anxiety/Depression. The first two subgroups were used for training and testing of the deep learning algorithm, and the third subgroup was used for external validation of the resulting model. Hypnograms were constructed via automatic sleep staging, with the 12-channel PSG data being transformed into three-channel RGB (red, green, blue channels) images for analysis. Composite patient images were generated and utilized for training the Xception model, which provided a validation set accuracy of 0.9782 on the ninth training epoch. In the independent test set, the model achieved a high accuracy (0.9688), precision (0.9533), recall (0.9630), and F1-score (0.9581). Classification performance of most other mainstream deep learning models was comparable. These findings suggest that machine learning techniques have the potential to accurately detect the presence of anxiety and depression from analysis of sleep study data. Further studies are needed to explore the utility of these techniques in the field of psychiatry.

Fast Speckle Noise Suppression Algorithm in Breast Ultrasound Image Using Three-Dimensional Deep Learning

The rapid development of ultrasound medical imaging technology has greatly broadened the scope of application of ultrasound, which has been widely used in the screening, diagnosis of breast diseases and so on. However, the presence of excessive speckle noise in breast ultrasound images can greatly reduce the image resolution and affect the observation and judgment of patients’ condition. Therefore, it is particularly important to investigate image speckle noise suppression. In the paper, we propose fast speckle noise suppression algorithm in breast ultrasound image using three-dimensional (3D) deep learning. Firstly, according to the gray value of the breast ultrasound image, the input breast ultrasound image contrast is enhanced using logarithmic and exponential transforms, and guided filter algorithm was used to enhance the details of glandular ultrasound image, and spatial high-pass filtering algorithm was used to suppress the excessive sharpening of breast ultrasound image to complete the pre-processing of breast ultrasound image and improve the image clarity; Secondly, the pre-processed breast ultrasound images were input into the 3D convolutional cloud neural network image speckle noise suppression model; Finally, the edge sensitive terms were introduced into the 3D convolutional cloud neural network to suppress the speckle noise of breast ultrasound images while retaining image edge information. The experiments demonstrate that the mean square error and false recognition rate all reduced to below 1.2% at the 100th iteration of training, and the 3D convolutional cloud neural network is well trained, and the signal-to-noise ratio of ultrasound image speckle noise suppression is greater than 60 dB, the peak signal-to-noise ratio is greater than 65 dB, the edge preservation index value exceeds the experimental threshold of 0.45, the speckle noise suppression time is low, the edge information is well preserved, and the image details are clearly visible. The speckle noise suppression time is low, the edge information is well preserved, and the image details are clearly visible, which can be applied to the field of breast ultrasound diagnosis.

Automatic sleep staging of EEG signals: recent development, challenges, and future directions

Modern deep learning holds a great potential to transform clinical practice on human sleep. Teaching a machine to carry out routine tasks would be a tremendous reduction in workload for clinicians. Sleep staging, a fundamental step in sleep practice, is a suitable task for this and will be the focus in this article. Recently, automatic sleep staging systems have been trained to mimic manual scoring, leading to similar performance to human sleep experts, at least on scoring of healthy subjects. Despite tremendous progress, we have not seen automatic sleep scoring adopted widely in clinical environments. This review aims to give a shared view of the authors on the most recent state-of-the-art development in automatic sleep staging, the challenges that still need to be addressed, and the future directions for automatic sleep scoring to achieve clinical value.

Transfer learning for non-image data in clinical research: A scoping review

BACKGROUND

Transfer learning is a form of machine learning where a pre-trained model trained on a specific task is reused as a starting point and tailored to another task in a different dataset. While transfer learning has garnered considerable attention in medical image analysis, its use for clinical non-image data is not well studied. Therefore, the objective of this scoping review was to explore the use of transfer learning for non-image data in the clinical literature.

METHODS AND FINDINGS

We systematically searched medical databases (PubMed, EMBASE, CINAHL) for peer-reviewed clinical studies that used transfer learning on human non-image data. We included 83 studies in the review. More than half of the studies (63%) were published within 12 months of the search. Transfer learning was most often applied to time series data (61%), followed by tabular data (18%), audio (12%) and text (8%). Thirty-three (40%) studies applied an image-based model to non-image data after transforming data into images (e.g. spectrograms). Twenty-nine (35%) studies did not have any authors with a health-related affiliation. Many studies used publicly available datasets (66%) and models (49%), but fewer shared their code (27%).

CONCLUSIONS

In this scoping review, we have described current trends in the use of transfer learning for non-image data in the clinical literature. We found that the use of transfer learning has grown rapidly within the last few years. We have identified studies and demonstrated the potential of transfer learning in clinical research in a wide range of medical specialties. More interdisciplinary collaborations and the wider adaption of reproducible research principles are needed to increase the impact of transfer learning in clinical research.

Automated Scoring of Sleep and Associated Events

Conventionally, sleep and associated events are scored visually by trained technologists according to the rules summarized in the American Academy of Sleep Medicine Manual. Since its first publication in 2007, the manual was continuously updated; the most recent version as of this writing was published in 2020. Human expert scoring is considered as gold standard, even though there is increasing evidence of limited interrater reliability between human scorers. Significant advances in machine learning have resulted in powerful methods for addressing complex classification problems such as automated scoring of sleep and associated events. Evidence is increasing that these autoscoring systems deliver performance comparable to manual scoring and offer several advantages to visual scoring: (1) avoidance of the rather expensive, time-consuming, and difficult visual scoring task that can be performed only by well-trained and experienced human scorers, (2) attainment of consistent scoring results, and (3) proposition of added value such as scoring in real time, sleep stage probabilities per epoch (hypnodensity), estimates of signal quality and sleep/wake-related features, identifications of periods with clinically relevant ambiguities (confidence trends), configurable sensitivity and rule settings, as well as cardiorespiratory sleep staging for home sleep apnea testing. This chapter describes the development of autoscoring systems since the first attempts in the 1970s up to the most recent solutions based on deep neural network approaches which achieve an accuracy that allows to use the autoscoring results directly for review and interpretation by a physician.

Air pollution and cardiorespiratory hospitalization, predictive modeling, and analysis using artificial intelligence techniques

Air pollution has a serious and adverse effect on human health, and it has become a risk to human welfare and health throughout the globe. One of the major effects of air pollution on health is hospitalizations associated with air pollution. Recently, the estimation and prediction of air pollution-based hospitalization is carried out using artificial intelligence (AI) and machine learning (ML) techniques, i.e., deep learning and long short-term memory (LSTM). However, there is ample room for improvement in the available applied methodologies to estimate and predict air pollution-based hospital admissions. In this paper, we present the modeling and analysis of air pollution and cardiorespiratory hospitalization. This study aims to investigate the association between cardiorespiratory hospitalization and air pollution, and predict cardiorespiratory hospitalization based on air pollution using the artificial intelligence (AI) techniques. We propose the enhanced long short-term memory (ELSTM) model and provide a comparison with other AI techniques, i.e., LSTM, DL, and vector autoregressive (VAR). This study was conducted at seven study locations in Klang Valley, Malaysia. The utilized dataset contains the data from January 2006 to December 2016 for five study locations, i.e., Klang (KLN), Shah Alam (SA), Putrajaya (PUJ), Petaling Jaya (PJ), and Cheras, Kuala Lumpur (CKL). The dataset for Banting contains data from April 2010 to December 2016, and the data for Batu Muda, Kuala Lumpur, contains data from January 2009 to December 2016. The prediction results show that the ELSTM model performed significantly better than other models in all study locations, with the best RMSE scores in Klang study location (ELSTM: 0.002, LSTM: 0.013, DL: 0.006, VAR: 0.066). The results also indicated that the proposed ELSTM model was able to detect and predict the trends of monthly hospitalization significantly better than the LSTM and other models in the study. Hence, we can conclude that we can utilize AI techniques to accurately predict cardiorespiratory hospitalization based on air pollution in Klang Valley, Malaysia.

Transferring Structured Knowledge in Unsupervised Domain Adaptation of a Sleep Staging Network

Automatic sleep staging based on deep learning (DL) has been attracting attention for analyzing sleep quality and determining treatment effects. It is challenging to acquire long-term sleep data from numerous subjects and manually labeling them even though most DL-based models are trained using large-scale sleep data to provide state-of-the-art performance. One way to overcome this data shortage is to create a pre-trained network with an existing large-scale dataset (source domain) that is applicable to small cohorts of datasets (target domain); however, discrepancies in data distribution between the domains prevent successful refinement of this approach. In this paper, we propose an unsupervised domain adaptation method for sleep staging networks to reduce discrepancies by realigning the domains in the same space and producing domain-invariant features. Specifically, in addition to a classical domain discriminator, we introduce local dis-criminators-subject and stage-to maintain the intrinsic structure of sleep data to decrease local misalignments while using adversarial learning to play a minimax game between the feature extractor and discriminators. Moreover, we present several optimization schemes during training because the conventional adversarial learning is not effective to our training scheme. We evaluate the performance of the proposed method by examining the staging performances of a baseline network compared with direct transfer (DT) learning in various conditions. The experimental results demonstrate that the proposed domain adaptation significantly improves the performance though it needs no labeled sleep data in target domain.

Invasive Electrophysiology for Circuit Discovery and Study of Comorbid Psychiatric Disorders in Patients With Epilepsy: Challenges, Opportunities, and Novel Technologies

Intracranial electroencephalographic (iEEG) recordings from patients with epilepsy provide distinct opportunities and novel data for the study of co-occurring psychiatric disorders. Comorbid psychiatric disorders are very common in drug-resistant epilepsy and their added complexity warrants careful consideration. In this review, we first discuss psychiatric comorbidities and symptoms in patients with epilepsy. We describe how epilepsy can potentially impact patient presentation and how these factors can be addressed in the experimental designs of studies focused on the electrophysiologic correlates of mood. Second, we review emerging technologies to integrate long-term iEEG recording with dense behavioral tracking in naturalistic environments. Third, we explore questions on how best to address the intersection between epilepsy and psychiatric comorbidities. Advances in ambulatory iEEG and long-term behavioral monitoring technologies will be instrumental in studying the intersection of seizures, epilepsy, psychiatric comorbidities, and their underlying circuitry.

Interrater sleep stage scoring reliability between manual scoring from two European sleep centers and automatic scoring performed by the artificial intelligence-based Stanford-STAGES algorithm

STUDY OBJECTIVES

The objective of this study was to evaluate interrater reliability between manual sleep stage scoring performed in 2 European sleep centers and automatic sleep stage scoring performed by the previously validated artificial intelligence-based Stanford-STAGES algorithm.

METHODS

Full night polysomnographies of 1,066 participants were included. Sleep stages were manually scored in Berlin and Innsbruck sleep centers and automatically scored with the Stanford-STAGES algorithm. For each participant, we compared (1) Innsbruck to Berlin scorings (INN vs BER); (2) Innsbruck to automatic scorings (INN vs AUTO); (3) Berlin to automatic scorings (BER vs AUTO); (4) epochs where scorers from Innsbruck and Berlin had consensus to automatic scoring (CONS vs AUTO); and (5) both Innsbruck and Berlin manual scorings (MAN) to the automatic ones (MAN vs AUTO). Interrater reliability was evaluated with several measures, including overall and sleep stage-specific Cohen's κ.

RESULTS

Overall agreement across participants was substantial for INN vs BER (κ = 0.66 ± 0.13), INN vs AUTO (κ = 0.68 ± 0.14), CONS vs AUTO (κ = 0.73 ± 0.14), and MAN vs AUTO (κ = 0.61 ± 0.14), and moderate for BER vs AUTO (κ = 0.55 ± 0.15). Human scorers had the highest disagreement for N1 sleep (κ_N1 = 0.40 ± 0.16 for INN vs BER). Automatic scoring had lowest agreement with manual scorings for N1 and N3 sleep (κ_N1 = 0.25 ± 0.14 and κ_N3 = 0.42 ± 0.32 for MAN vs AUTO).

CONCLUSIONS

Interrater reliability for sleep stage scoring between human scorers was in line with previous findings, and the algorithm achieved an overall substantial agreement with manual scoring. In this cohort, the Stanford-STAGES algorithm showed similar performances to the ones achieved in the original study, suggesting that it is generalizable to new cohorts. Before its integration in clinical practice, future independent studies should further evaluate it in other cohorts.

Can Big Data guide prognosis and clinical decisions in epilepsy?

Big Data is no longer a novel concept in health care. Its promise of positive impact is not only undiminished, but daily enhanced by seemingly endless possibilities. Epilepsy is a disorder with wide heterogeneity in both clinical and research domains, and thus lends itself to Big Data concepts and techniques. It is therefore inevitable that Big Data will enable multimodal research, integrating various aspects of "-omics" domains, such as phenome, genome, microbiome, metabolome, and proteome. This scope and granularity have the potential to change our understanding of prognosis and mortality in epilepsy. The scale of new discovery is unprecedented due to the possibilities promised by advances in machine learning, in particular deep learning. The subsequent possibilities of personalized patient care through clinical decision support systems that are evidence-based, adaptive, and iterative seem to be within reach. A major objective is not only to inform decision-making, but also to reduce uncertainty in outcomes. Although the adoption of electronic health record (EHR) systems is near universal in the United States, for example, advanced clinical decision support in or ancillary to EHRs remains sporadic. In this review, we discuss the role of Big Data in the development of clinical decision support systems for epilepsy care, prognostication, and discovery.

A Systematic Review of Sensing Technologies for Wearable Sleep Staging

Designing wearable systems for sleep detection and staging is extremely challenging due to the numerous constraints associated with sensing, usability, accuracy, and regulatory requirements. Several researchers have explored the use of signals from a subset of sensors that are used in polysomnography (PSG), whereas others have demonstrated the feasibility of using alternative sensing modalities. In this paper, a systematic review of the different sensing modalities that have been used for wearable sleep staging is presented. Based on a review of 90 papers, 13 different sensing modalities are identified. Each sensing modality is explored to identify signals that can be obtained from it, the sleep stages that can be reliably identified, the classification accuracy of systems and methods using the sensing modality, as well as the usability constraints of the sensor in a wearable system. It concludes that the two most common sensing modalities in use are those based on electroencephalography (EEG) and photoplethysmography (PPG). EEG-based systems are the most accurate, with EEG being the only sensing modality capable of identifying all the stages of sleep. PPG-based systems are much simpler to use and better suited for wearable monitoring but are unable to identify all the sleep stages.

Machine and Deep Learning in Molecular and Genetic Aspects of Sleep Research

Epidemiological sleep research strives to identify the interactions and causal mechanisms by which sleep affects human health, and to design intervention strategies for improving sleep throughout the lifespan. These goals can be advanced by further focusing on the environmental and genetic etiology of sleep disorders, and by development of risk stratification algorithms, to identify people who are at risk or are affected by, sleep disorders. These studies rely on comprehensive sleep-related data which often contains complex multi-dimensional physiological and molecular measurements across multiple timepoints. Thus, sleep research is well-suited for the application of computational approaches that can handle high-dimensional data. Here, we survey recent advances in machine and deep learning together with the availability of large human cohort studies with sleep data that can jointly drive the next breakthroughs in the sleep-research field. We describe sleep-related data types and datasets, and present some of the tasks in the field that can be targets for algorithmic approaches, as well as the challenges and opportunities in pursuing them.