RESUMEN

Automatic sleep staging has been an active field of development. Despite multiple efforts, the area remains a focus of research interest. Indeed, while promising results have reported in past literature, uptake of automatic sleep scoring in the clinical setting remains low. One of the current issues regards the difficulty to generalization performance results beyond the local testing scenario, i.e. across data from different clinics. Issues derived from data-privacy restrictions, that generally apply in the medical domain, pose additional difficulties in the successful development of these methods. We propose the use of several decentralized deep-learning approaches, namely ensemble models and federated learning, for robust inter-database performance generalization and data-privacy preservation in automatic sleep staging scenario. Specifically, we explore four ensemble combination strategies (max-voting, output averaging, size-proportional weighting, and Nelder-Mead) and present a new federated learning algorithm, so-called sub-sampled federated stochastic gradient descent (ssFedSGD). To evaluate generalization capabilities of such approaches, experimental procedures are carried out using a leaving-one-database-out direct-transfer scenario on six independent and heterogeneous public sleep staging databases. The resulting performance is compared with respect to two baseline approaches involving single-database and centralized multiple-database derived models. Our results show that proposed decentralized learning methods outperform baseline local approaches, and provide similar generalization results to centralized database-combined approaches. We conclude that these methods are more preferable choices, as they come with additional advantages concerning improved scalability, flexible design, and data-privacy preservation.

Asunto(s)

RESUMEN

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.

Asunto(s)

RESUMEN

BACKGROUND: Polysomnography is the gold standard for measuring and detecting sleep patterns. In recent years, activity wristbands have become popular because they record continuous data in real time. Hence, comprehensive validation studies are needed to analyze the performance and reliability of these devices in the recording of sleep parameters. OBJECTIVE: This study compared the performance of one of the best-selling activity wristbands, the Xiaomi Mi Band 5, against polysomnography in measuring sleep stages. METHODS: This study was carried out at a hospital in A Coruña, Spain. People who were participating in a polysomnography study at a sleep unit were recruited to wear a Xiaomi Mi Band 5 simultaneously for 1 night. The total sample consisted of 45 adults, 25 (56%) with sleep disorders (SDis) and 20 (44%) without SDis. RESULTS: Overall, the Xiaomi Mi Band 5 displayed 78% accuracy, 89% sensitivity, 35% specificity, and a Cohen κ value of 0.22. It significantly overestimated polysomnography total sleep time (P=.09), light sleep (N1+N2 stages of non-rapid eye movement [REM] sleep; P=.005), and deep sleep (N3 stage of non-REM sleep; P=.01). In addition, it underestimated polysomnography wake after sleep onset and REM sleep. Moreover, the Xiaomi Mi Band 5 performed better in people without sleep problems than in those with sleep problems, specifically in detecting total sleep time and deep sleep. CONCLUSIONS: The Xiaomi Mi Band 5 can be potentially used to monitor sleep and to detect changes in sleep patterns, especially for people without sleep problems. However, additional studies are necessary with this activity wristband in people with different types of SDis. TRIAL REGISTRATION: ClinicalTrials.gov NCT04568408; https://clinicaltrials.gov/ct2/show/NCT04568408. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.3390/ijerph18031106.

Asunto(s)

RESUMEN

STUDY OBJECTIVES: To investigate inter-scorer agreement and scoring time differences associated with visual and computer-assisted analysis of polysomnographic (PSG) recordings. METHODS: A group of 12 expert scorers reviewed 5 PSGs that were independently selected in the context of each of the following tasks: (i) sleep staging, (ii) scoring of leg movements, (iii) detection of respiratory (apneic-related) events, and (iv) of electroencephalographic (EEG) arousals. All scorers independently reviewed the same recordings, hence resulting in 20 scoring exercises per scorer from an equal amount of different subjects. The procedure was repeated, separately, using the classical visual manual approach and a computer-assisted (semi-automatic) procedure. Resulting inter-scorer agreement and scoring times were examined and compared among the two methods. RESULTS: Computer-assisted sleep scoring showed a consistent and statistically relevant effect toward less time required for the completion of each of the PSG scoring tasks. Gain factors ranged from 1.26 (EEG arousals) to 2.41 (leg movements). Inter-scorer kappa agreement was also consistently increased with the use of supervised semi-automatic scoring. Specifically, agreement increased from Κ = 0.76 to K = 0.80 (sleep stages), Κ = 0.72 to K = 0.91 (leg movements), Κ = 0.55 to K = 0.66 (respiratory events), and Κ = 0.58 to Κ = 0.65 (EEG arousals). Inter-scorer agreement on the examined set of diagnostic indices did also show a trend toward higher Interclass Correlation Coefficient scores when using the semi-automatic scoring approach. CONCLUSIONS: Computer-assisted analysis can improve inter-scorer agreement and scoring times associated with the review of PSG studies resulting in higher efficiency and overall quality in the diagnosis sleep disorders.

Asunto(s)

RESUMEN

STUDY OBJECTIVES: Development of inter-database generalizable sleep staging algorithms represents a challenge due to increased data variability across different datasets. Sharing data between different centers is also a problem due to potential restrictions due to patient privacy protection. In this work, we describe a new deep learning approach for automatic sleep staging, and address its generalization capabilities on a wide range of public sleep staging databases. We also examine the suitability of a novel approach that uses an ensemble of individual local models and evaluate its impact on the resulting inter-database generalization performance. METHODS: A general deep learning network architecture for automatic sleep staging is presented. Different preprocessing and architectural variant options are tested. The resulting prediction capabilities are evaluated and compared on a heterogeneous collection of six public sleep staging datasets. Validation is carried out in the context of independent local and external dataset generalization scenarios. RESULTS: Best results were achieved using the CNN_LSTM_5 neural network variant. Average prediction capabilities on independent local testing sets achieved 0.80 kappa score. When individual local models predict data from external datasets, average kappa score decreases to 0.54. Using the proposed ensemble-based approach, average kappa performance on the external dataset prediction scenario increases to 0.62. To our knowledge this is the largest study by the number of datasets so far on validating the generalization capabilities of an automatic sleep staging algorithm using external databases. CONCLUSIONS: Validation results show good general performance of our method, as compared with the expected levels of human agreement, as well as to state-of-the-art automatic sleep staging methods. The proposed ensemble-based approach enables flexible and scalable design, allowing dynamic integration of local models into the final ensemble, preserving data locality, and increasing generalization capabilities of the resulting system at the same time.

Asunto(s)

RESUMEN

STUDY OBJECTIVES: To investigate (1) the effect of different scoring rules on leg movement (LM) classification in patients with obstructive sleep apnea (OSA); (2) determinants of respiratory event related leg movements (rLM); and (3) to relate LM parameters to clinical outcomes. METHODS: (1) LM classification was compared between the World Association of Sleep Medicine (WASM) 2006 and the WASM 2016 rules in 336 participants with apnea hypopnea index (AHI) ≥ 5; (2) determinants and features of rLM were investigated with logistic mixed regression in 172 participants with AHI ≥ 10 and respiratory disturbance index (RDI) ≥ 15, and (3) LM parameters were compared for patients with and without cardiovascular events and related to continuous positive airway pressure (CPAP) adherence. RESULTS: WASM-2016 scoring significantly reduced periodic limb movements of sleep (PLMS) frequency in OSA participants even when only considering the new periodicity criteria. Probability of rLM was strongly increased when respiratory events ended with an arousal, but rLM probability was lower for hypopneas and respiratory effort-related sleep arousal (RERAs) than for obstructive apneas. In participants with frequent non-respiratory PLMS, rLM were more frequent and behaved more PLMS-like. In participants without PLMS, rLM probability mostly depended on respiratory event features. LM parameters were neither related to cardiovascular event risk nor to CPAP-adherence. CONCLUSIONS: It is likely that the PLMS frequency in OSA populations has been previously overestimated. Our results suggest that there are two types of rLM, true periodic ones that happen to synchronize with the respiratory events, and periodic appearing but respiratory driven LM, and that the presence of non-respiratory PLMS is instrumental in distinguishing between the two.

Asunto(s)

RESUMEN

In this work we examine some of the problems associated with the development of machine learning models with the objective to achieve robust generalization capabilities on common-task multiple-database scenarios. Referred to as the "database variability problem", we focus on a specific medical domain (sleep staging in sleep medicine) to show the non-triviality of translating the estimated model's local generalization capabilities into independent external databases. We analyze some of the scalability problems when multiple-database data are used as inputs to train a single learning model. Then, we introduce a novel approach based on an ensemble of local models, and we show its advantages in terms of inter-database generalization performance and data scalability. In addition, we analyze different model configurations and data pre-processing techniques to determine their effects on the overall generalization performance. For this purpose, we carry out experimentation that involves several sleep databases and evaluates different machine learning models based on convolutional neural networks.

Asunto(s)

RESUMEN

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.

Asunto(s)

RESUMEN

BACKGROUND: Periodic leg movements during sleep (PLMS) have been associated with an increased risk for cardiovascular diseases and there is a high prevalence of PLMS found in patients with obstructive sleep apnea syndrome (OSAS). We evaluated patients with transient ischemic attack (TIA) for PLMS and respiratory related leg movements (RRLM), versus a control group without TIA. METHODS: Twenty-five patients with TIA and 34 patients with no vascular diagnosis were referred for polysomnography. Diagnosis of PLMS was made if the periodic leg movement index (PLMI) was ≥5 and clinical significant as PLMI ≥15. RESULTS: There was no significant difference in PLMI ≥5 and ≥15 between patients with and without TIA. In the absence of OSAS, 2 out of 5 TIA patients (40%) had a PLMI ≥15 compared to 1 of the 19 patients without TIA (5%; p = 0.037). There was no increase in RRLMs when OSAS was present. CONCLUSIONS: TIA patients did not have higher PLMI compared to controls, and in the presence of OSAS, there was no increase in RRLMs compared to patients without TIA. In selective patients, PLMS could be associated with cardiovascular diseases, since PLMS was clinically more often found in the TIA group without OSAS.

Asunto(s)

RESUMEN

Obstructive sleep apnea (OSA) is a known-risk factor for cardiovascular diseases. There are indications that treatment with continuous positive airway pressure (CPAP) reduces the risk of new cardiovascular events. In this study, we analyzed the incidence of cardiovascular events in patients with OSA and compared for the impact of CPAP therapy. All polysomnographies performed in 2009 and 2010 were selected with an AHI ≥5 and patients older than 18 years. These 1110 patients were approached with a questionnaire about cardiovascular events and CPAP treatment. Finally, 554 patients were included in analyses. CPAP treatment was based on compliance (level 1 treatment) and extended with residual respiratory events (level 2 treatment). OSA was set as AHI ≥5 and classified in mild (AHI 5-15), moderate (AHI 15-30) and severe (AHI ≥30) OSA. 50 cardiovascular events occurred in 44 patients during follow-up (mean follow-up time 5.9 years) in 554 patients. The events were significantly higher in patients with increasing classification of OSA-severity (p = 0.016). A first-ever cardiovascular event did not differ significantly between mild, moderate and severe OSA. Untreated CPAP patients had significantly more cardiovascular events as compared to treated patients with a hazard ratio of 2.66 partially adjusted for age, AHI and smoking. There was no significant contribution of other cardiovascular risk factors. Patients with OSA with an indication for CPAP treatment have more cardiovascular events when untreated compared to treated patients. This indicates that treatment of OSA by CPAP can reduce the risk for cardiovascular events.

Asunto(s)

RESUMEN

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.

Asunto(s)

RESUMEN

Some approaches have been published in the past using Heart Rate Variability (HRV) spectral features for the screening of Sleep Apnea-Hypopnea Syndrome (SAHS) patients. However there is a big variability among these methods regarding the selection of the source signal and the specific spectral components relevant to the analysis. In this study we investigate the use of the Heart Timing (HT) as the source signal in comparison to the classical approaches of Heart Rate (HR) and Heart Period (HP). This signal has the theoretical advantage of being optimal under the Integral Pulse Frequency Modulation (IPFM) model assumption. Only spectral bands defined as standard for the study of HRV are considered, and for each method the so-called LF/HF and VLFn features are derived. A comparative statistical analysis between the different resulting methods is performed, and subject classification is investigated by means of ROC analysis and a Naïve-Bayes classifier. The standard Apnea-ECG database is used for validation purposes. Our results show statistical differences between SAHS patients and controls for all the derived features. In the subject classification task the best performance in the testing set was obtained using the LF/HF ratio derived from the HR signal (Area under ROC curve=0.88). Only slight differences are obtained due to the effect of changing the source signal. The impact of using the HT signal in this domain is therefore limited, and has not shown relevant differences with respect to the use of the classical approaches of HR or HP.

Asunto(s)

RESUMEN

Automatic diagnosis of the Sleep Apnea-Hypopnea Syndrome (SAHS) has become an important area of research due to the growing interest in the field of sleep medicine and the costs associated with its manual diagnosis. The increment and heterogeneity of the different techniques, however, make it somewhat difficult to adequately follow the recent developments. A literature review within the area of computer-assisted diagnosis of SAHS has been performed comprising the last 15 years of research in the field. Screening approaches, methods for the detection and classification of respiratory events, comprehensive diagnostic systems, and an outline of current commercial approaches are reviewed. An overview of the different methods is presented together with validation analysis and critical discussion of the current state of the art.

RESUMEN

Automatic diagnosis of the Sleep Apnea-Hypopnea Syndrome (SAHS) has become an important area of research due to the growing interest in the field of sleep medicine, and the costs associated to its manual diagnosis. The increment and heterogeneity of the different techniques, however, makes somewhat difficult to adequately follow recent developments. In this paper an overview within the area of computer-assisted diagnosis of SAHS has been performed. This overview of the different methods is presented together with a critical discussion of the current state-of-the-art.

Asunto(s)

RESUMEN

This paper presents a comparative study over the respiratory pattern classification task involving three missing data imputation techniques, and four different machine learning algorithms. The main goal was to find a classifier that achieves the best accuracy results using a scalable imputation method in comparison to the method used in a previous work of the authors. The results obtained show that the Self-organization maps imputation method allows any classifier to achieve improvements over the rest of the imputation methods, and that the Feedforward neural network classifier offers the best performance regardless the imputation method used.

Asunto(s)

RESUMEN

This work deals with the development of an intelligent approach for clinical decision making in the diagnosis of the Sleep Apnea/Hypopnea Syndrome, SAHS, from the analysis of respiratory signals and oxygen saturation in arterial blood, SaO2. In order to accomplish the task the proposed approach makes use of different artificial intelligence techniques and reasoning processes being able to deal with imprecise data. These reasoning processes are based on fuzzy logic and on temporal analysis of the information. The developed approach also takes into account the possibility of artifacts in the monitored signals. Detection and characterization of signal artifacts allows detection of false positives. Identification of relevant diagnostic patterns and temporal correlation of events is performed through the implementation of temporal constraints.