Sleep quality prediction in caregivers using physiological signals

Can Sleep Quality Attributes be Predicted from Physical Activity in Everyday Settings?

Sleep is crucial for physical, mental, and emotional well-being. Physical activity and sleep are known to be interrelated; however, limited research has been performed to investigate their interactions in long-term. Conventional studies have presented sleep quality prediction, focusing on a single sleep quality aspect, such as sleep efficiency. In addition, the relationship between daily physical activity and sleep quality has yet to be explored, despite physical activities being utilized in previous studies for sleep quality prediction. In this paper, we develop an Extreme Gradient boosting method to predict sleep duration, sleep efficiency, and deep sleep based on users' daily activity information collected from wearable devices. Our model is trained and tested using data collected with an OURA ring from 34 pregnant mothers for six months under free-living conditions. Our finding shows an accuracy of 90.58%, 95.38%, and 91.45% for sleep duration, efficiency, and deep sleep, respectively. Moreover, we assess the contribution of each physical activity parameter to the prediction results using the Shapley Additive Explanations method. Our results indicate that sedentary time is the most influential parameter for sleep duration prediction, while the inactive time feature (e.g., resting or lying down) has a strong negative relationship with sleep efficiency, and the pregnancy week is the most critical parameter for deep sleep prediction.

Interactive slide selection algorithm and machine learning in psychophysiological memory testing

Objective. To present a new type of concealed information test (CIT) that implements the interactive slide selection (ISS) algorithm and compare its effectiveness with a standard CIT (sCIT).Approach. The ISS algorithm presents slides interactively, based on the analysis of electrodermal activity, while sCIT presents slides in a predefined, sequential order. The algorithm automatically selects irrelevant, relevant, and control slides and presents them at the moment which is physiologically most suitable for electrodermal response detection. To compare the ISS-based CIT (issCIT) and sCIT, two objects, a bag, and a wallet, were presented to 64 participants, 32 of whomwere analyzed with sCIT, and another 32 with issCIT.Main results. The results show that ISS had significantly better true/false predictions (Fisher's exact test,p< 0.01). Also, the number of false positives (FPs) was significantly lower in the issCIT group in comparison with sCIT (Fisher's exact test,p< 0.001). Machine learning (ML) classifiers improved precision from 49% to 79% in the sCIT group (McNemar's test,p< 0.05), and from 85% to 100% in the issCIT group (McNemar's test,p< 0.05). The testing time in the issCIT group ranged between 42 and 107 s, while the average was 53 s. In the sCIT group, the testing time was always 330 s.Significance. Under the presented experimental settings, the ISS algorithm obtained significantly better classification results compared to sCIT, while the application of the ML algorithms managed to improve the classification results in both groups reaching a precision of 100%. The ISS algorithm allowed for a much shorter testing time compared to sCIT.

Analysis of Sleep Quality in People With Dementia: A Preliminary Study

Background: Between 25% and 60% of subjects with dementia have shown sleep disturbances. Causes are diverse and seem to be related to factors such as aging, the presence of psychiatric diseases, or the consumption of drugs. This study aims to determine the influence of dementia on sleep quality and to analyze the factors that influence sleep quality in subjects ≥65 years. Methods: Thirty-one subjects were studied (15 living with dementia). PSQI was administered and statistical analysis compared the results among categories of other variables (age, gender, coffee consumption, drugs, BMI, psychiatric diseases). This study took place in Spain. Results: A prevalence of 46.7% of sleep disturbances was found in subjects with dementia. No significant differences were observed in the total score obtained in the PSQI between the dementia group (6.06 ± 3.78 points) and the group without dementia (7 ± 5.65 points). A significant inverse relationship was found between the sleep quality and the number of daily drugs and the presence of psychiatric diseases. Conclusion: Poor sleep quality affects people with dementia, however, we cannot affirm that dementia is the cause of it. Consumption of daily drugs and psychiatric diseases are factors that influence the sleep quality in subjects aged ≥65 years.

Feasibility of wearable devices and machine learning for sleep classification in children with Rett syndrome: A pilot study

Sleep is vital to many processes involved in the well-being and health of children; however, it is estimated that 80% of children with Rett syndrome suffer from sleep disorders. Caregiver reports and questionnaires, which are the current method of studying sleep, are prone to observer bias and missed information. Polysomnography is considered the gold standard for sleep analysis but is labor and cost-intensive and limits the frequency of data collection for sleep disorder studies. Wearable digital health technologies, such as actigraphy devices, have shown potential and feasibility as a method for sleep analysis in Rett syndrome, but have not been validated against polysomnography. Furthermore, the collected accelerometer data has limitations due to the rigidity, periodic limb movement, and involuntary muscle contractions prevalent in Rett syndrome. Heart rate and electrodermal activity, along with other physiological signals, have been linked to sleep stages and can be utilized with machine learning to provide better resistance to noise and false positives than actigraphy. This research aims to address the gap in Rett syndrome sleep analysis by comparing the performance of a machine learning model utilizing both accelerometer data and physiological data features to the gold-standard polysomnography for sleep analysis in Rett syndrome. Our analytical validation pilot study (n = 7) found that using physiological and accelerometer features, our machine learning models can differentiate between awake, non-rapid eye movement sleep, and rapid eye movement sleep in Rett syndrome children with an accuracy of 85.1% when using an individual model. Additionally, this work demonstrates that it is feasible to use digital health technologies in Rett syndrome, even at a young age, without data loss or interference from repetitive movements that are characteristic of Rett syndrome.

Machine Learning Techniques for Arousal Classification from Electrodermal Activity: A Systematic Review

This article introduces a systematic review on arousal classification based on electrodermal activity (EDA) and machine learning (ML). From a first set of 284 articles searched for in six scientific databases, fifty-nine were finally selected according to various criteria established. The systematic review has made it possible to analyse all the steps to which the EDA signals are subjected: acquisition, pre-processing, processing and feature extraction. Finally, all ML techniques applied to the features of these signals for arousal classification have been studied. It has been found that support vector machines and artificial neural networks stand out within the supervised learning methods given their high-performance values. In contrast, it has been shown that unsupervised learning is not present in the detection of arousal through EDA. This systematic review concludes that the use of EDA for the detection of arousal is widely spread, with particularly good results in classification with the ML methods found.

A multidevice and multimodal dataset for human energy expenditure estimation using wearable devices

We present a multi-device and multi-modal dataset, called WEEE, collected from 17 participants while they were performing different physical activities. WEEE contains: (1) sensor data collected using seven wearable devices placed on four body locations (head, ear, chest, and wrist); (2) respiratory data collected with an indirect calorimeter serving as ground-truth information; (3) demographics and body composition data (e.g., fat percentage); (4) intensity level and type of physical activities, along with their corresponding metabolic equivalent of task (MET) values; and (5) answers to questionnaires about participants' physical activity level, diet, stress and sleep. Thanks to the diversity of sensors and body locations, we believe that the dataset will enable the development of novel human energy expenditure (EE) estimation techniques for a diverse set of application scenarios. EE refers to the amount of energy an individual uses to maintain body functions and as a result of physical activity. A reliable estimate of people's EE thus enables computing systems to make inferences about users' physical activity and help them promoting a healthier lifestyle.

Prediction of sleep quality among university students after analyzing lifestyles, sports habits, and mental health

The aim of this study was to develop and validate a prediction model to evaluate the risk of poor sleep quality. We performed a cross-sectional study and enrolled 1,928 college students from five universities between September and November 2021. The quality of sleep was evaluated using the Chinese version of the Pittsburgh Sleep Quality Index (PSQI). Participants were divided into a training (n = 1,555) group and a validation (n = 373) group. The training group was used to establish the model, and the validation group was used to validate the predictive effectiveness of the model. The risk classification of all participants was performed based on the optimal threshold of the model. Of all enrolled participants, 45.07% (869/1,928) had poor sleep quality (PSQI score ≧ 6 points). Multivariate analysis showed that factors such as older age, a higher grade, previous smoking, drinking, midday rest, chronic disease, anxiety, and stress were significantly associated with a higher rate of poor sleep quality, while preference for vegetables was significantly associated with better sleep quality, and all these variables were included to develop the prediction model. The area under the curve (AUC) was 0.765 [95% confidence interval (CI): 0.742-0.789] in the training group and 0.715 (95% CI: 0.664-0.766) in the validation group. Corresponding discrimination slopes were 0.207 and 0.167, respectively, and Brier scores were 0.195 and 0.221, respectively. Calibration curves showed favorable matched consistency between the predicted and actual probability of poor sleep quality in both groups. Based on the optimal threshold, the actual probability of poor sleep quality was 29.03% (317/1,092) in the low-risk group and 66.03% (552/836) in the high-risk group (P < 0.001). A nomogram was presented to calculate the probability of poor sleep quality to promote the applicationof the model. The prediction model can be a helpful tool to stratify sleep quality, especially among university students. Some intervention measures or preventive strategies to quit smoking and drinking, eat more vegetables, avoid midday rest, treat chronic disease, and alleviate anxiety and stress may be considerably beneficial in improving sleep quality.

Wavelet-Based Biphase Analysis of Brain Rhythms in Automated Wake-Sleep Classification

Many studies in the field of sleep have focused on connectivity and coherence. Still, the nonstationary nature of electroencephalography (EEG) makes many of the previous methods unsuitable for automatic sleep detection. Time-frequency representations and high-order spectra are applied to nonstationary signal analysis and nonlinearity investigation, respectively. Therefore, combining wavelet and bispectrum, wavelet-based bi-phase (Wbiph) was proposed and used as a novel feature for sleep-wake classification. The results of the statistical analysis with emphasis on the importance of the gamma rhythm in sleep detection show that the Wbiph is more potent than coherence in the wake-sleep classification. The Wbiph has not been used in sleep studies before. However, the results and inherent advantages, such as the use of wavelet and bispectrum in its definition, suggest it as an excellent alternative to coherence. In the next part of this paper, a convolutional neural network (CNN) classifier was applied for the sleep-wake classification by Wbiph. The classification accuracy was 97.17% in nonLOSO and 95.48% in LOSO cross-validation, which is the best among previous studies on sleep-wake classification.

Current Challenges and Future Opportunities for XAI in Machine Learning-Based Clinical Decision Support Systems: A Systematic Review

Machine Learning and Artificial Intelligence (AI) more broadly have great immediate and future potential for transforming almost all aspects of medicine. However, in many applications, even outside medicine, a lack of transparency in AI applications has become increasingly problematic. This is particularly pronounced where users need to interpret the output of AI systems. Explainable AI (XAI) provides a rationale that allows users to understand why a system has produced a given output. The output can then be interpreted within a given context. One area that is in great need of XAI is that of Clinical Decision Support Systems (CDSSs). These systems support medical practitioners in their clinic decision-making and in the absence of explainability may lead to issues of under or over-reliance. Providing explanations for how recommendations are arrived at will allow practitioners to make more nuanced, and in some cases, life-saving decisions. The need for XAI in CDSS, and the medical field in general, is amplified by the need for ethical and fair decision-making and the fact that AI trained with historical data can be a reinforcement agent of historical actions and biases that should be uncovered. We performed a systematic literature review of work to-date in the application of XAI in CDSS. Tabular data processing XAI-enabled systems are the most common, while XAI-enabled CDSS for text analysis are the least common in literature. There is more interest in developers for the provision of local explanations, while there was almost a balance between post-hoc and ante-hoc explanations, as well as between model-specific and model-agnostic techniques. Studies reported benefits of the use of XAI such as the fact that it could enhance decision confidence for clinicians, or generate the hypothesis about causality, which ultimately leads to increased trustworthiness and acceptability of the system and potential for its incorporation in the clinical workflow. However, we found an overall distinct lack of application of XAI in the context of CDSS and, in particular, a lack of user studies exploring the needs of clinicians. We propose some guidelines for the implementation of XAI in CDSS and explore some opportunities, challenges, and future research needs.

Visualising the knowledge structure and evolution of wearable device research

In recent years, the literature associated with wearable devices has grown rapidly, but few studies have used bibliometrics and a visualisation approach to conduct deep mining and reveal a panorama of the wearable devices field. To explore the foundational knowledge and research hotspots of the wearable devices field, this study conducted a series of bibliometric analyses on the related literature, including papers' production trends in the field and the distribution of countries, a keyword co-occurrence analysis, theme evolution analysis and research hotspots and trends for the future. By conducting a literature content analysis and structure analysis, we found the following: (a) The subject evolution path includes sensor research, sensitivity research and multi-functional device research. (b) Wearable device research focuses on information collection, sensor materials, manufacturing technology and application, artificial intelligence technology application, energy supply and medical applications. The future development trend will be further studied in combination with big data analysis, telemedicine and personalised precision medical application.

Accuracy of Mobile Applications versus Wearable Devices in Long-Term Step Measurements

Fitness sensors and health systems are paving the way toward improving the quality of medical care by exploiting the benefits of new technology. For example, the great amount of patient-generated health data available today gives new opportunities to measure life parameters in real time and create a revolution in communication for professionals and patients. In this work, we concentrated on the basic parameter typically measured by fitness applications and devices-the number of steps taken daily. In particular, the main goal of this study was to compare the accuracy and precision of smartphone applications versus those of wearable devices to give users an idea about what can be expected regarding the relative difference in measurements achieved using different system typologies. In particular, the data obtained showed a difference of approximately 30%, proving that smartphone applications provide inaccurate measurements in long-term analysis, while wearable devices are precise and accurate. Accordingly, we challenge the reliability of previous studies reporting data collected with phone-based applications, and besides discussing the current limitations, we support the use of wearable devices for mHealth.

Predicting Sleep Quality in Osteoporosis Patients Using Electronic Health Records and Heart Rate Variability

Sleep quality (SQ) is one of the most well-known factors in daily work performance. Sleep is usually analyzed using polysomnography (PSG) by attaching electrodes to the bodies of participants, which is likely sleep destructive. As a result, investigating SQ using a more easy-to-use and cost-effective methodology is currently a hot topic. To avoid overfitting concerns, one likely methodology for predicting SQ can be achieved by reducing the number of utilized signals. In this paper, we propose three methodologies based on electronic health records and heart rate variability (HRV). To evaluate the performance of the proposed methods, several experiments have been conducted using the Osteoporotic Fractures in Men (MrOS) sleep dataset. The experimental results reveal that a deep neural network methodology can achieve an accuracy of 0.6 in predicting light, medium, and deep SQ using only ECG signals recorded during PSG. This outcome demonstrates the capability of using HRV features, which are effortlessly measurable by easy-to-use and cost-effective wearable devices, in predicting SQ.

Developing an ontology for representing the domain knowledge specific to non-pharmacological treatment for agitation in dementia

INTRODUCTION

A large volume of clinical care data has been generated for managing agitation in dementia. However, the valuable information in these data has not been used effectively to generate insights for improving the quality of care. Application of artificial intelligence technologies offers us enormous opportunities to reuse these data. For health data science to achieve this, this study focuses on using ontology to coding clinical knowledge for non-pharmacological treatment of agitation in a machine-readable format.

METHODS

The resultant ontology-Dementia-Related Agitation Non-Pharmacological Treatment Ontology (DRANPTO)-was developed using a method adopted from the NeOn methodology.

RESULTS

DRANPTO consisted of 569 concepts and 48 object properties. It meets the standards for biomedical ontology.

DISCUSSION

DRANPTO is the first comprehensive semantic representation of non-pharmacological management for agitation in dementia in the long-term care setting. As a knowledge base, it will play a vital role to facilitate the development of intelligent systems for managing agitation in dementia.