Background: Conventional daytime monitoring in a single day may be influenced by factors such as motion artifacts and emotions, and continuous monitoring of nighttime heart rate variability (HRV) and respiration to assist in chronic obstructive pulmonary disease (COPD) diagnosis has not been reported yet.
Objective: The aim of this study was to explore and compare the effects of continuously monitored HRV, heart rate (HR), and respiration during night sleep on the remote diagnosis of COPD.
Methods: We recruited patients with different severities of COPD and healthy controls between January 2021 and November 2022. Vital signs such as HRV, HR, and respiration were recorded using noncontact bed sensors from 10 PM to 8 AM of the following day, and the recordings of each patient lasted for at least 30 days. We obtained statistical means of HRV, HR, and respiration over time periods of 7, 14, and 30 days by continuous monitoring. Additionally, the effects that the statistical means of HRV, HR, and respiration had on COPD diagnosis were evaluated at different times of recordings.
Results: In this study, 146 individuals were enrolled: 37 patients with COPD in the case group and 109 participants in the control group. The median number of continuous night-sleep monitoring days per person was 56.5 (IQR 32.0-113.0) days. Using the features regarding the statistical means of HRV, HR, and respiration over 1, 7, 14, and 30 days, binary logistic regression classification of COPD yielded an accuracy, Youden index, and area under the receiver operating characteristic curve of 0.958, 0.904, and 0.989, respectively. The classification performance for COPD diagnosis was directionally proportional to the monitoring duration of vital signs at night. The importance of the features for diagnosis was determined by the statistical means of respiration, HRV, and HR, which followed the order of respiration > HRV > HR. Specifically, the statistical means of the duration of respiration rate faster than 21 times/min (RRF), high frequency band power of 0.15-0.40 Hz (HF), and respiration rate (RR) were identified as the top 3 most significant features for classification, corresponding to cutoff values of 0.1 minute, 1316.3 nU, and 16.3 times/min, respectively.
Conclusions: Continuous monitoring of nocturnal vital signs has significant potential for the remote diagnosis of COPD. As the duration of night-sleep monitoring increased from 1 to 30 days, the statistical means of HRV, HR, and respiration showed a better reflection of an individual's health condition compared to monitoring the vital signs in a single day or night, and better was the classification performance for COPD diagnosis. Further, the statistical means of RRF, HF, and RR are crucial features for diagnosing COPD, demonstrating the importance of monitoring HRV and respiration during night sleep.
Background: Cancer pain remains highly prevalent and persistent throughout survivorship, and it is crucial to investigate the potential of leveraging the advanced features of mobile health (mHealth) apps to empower individuals to self-manage their pain.
Objective: This review aims to comprehensively understand the acceptability, users' experiences, and effectiveness of mHealth apps in supporting cancer pain self-management.
Methods: We conducted an integrative review following Souza and Whittemore and Knafl's 6 review processes. Literature was searched in PubMed, Scopus, CINAHL Plus with Full Text, PsycINFO, and Embase, from 2013 to 2023. Keywords including "cancer patients," "pain," "self-management," "mHealth applications," and relevant synonyms were used in the search. The Johns Hopkins research evidence appraisal tool was used to evaluate the quality of eligible studies. A narrative synthesis was conducted to analyze the extracted data.
Results: A total of 20 studies were included, with the overall quality rated as high (n=15) to good (n=5). Using mHealth apps to monitor and manage pain was acceptable for most patients with cancer. The internal consistency of the mHealth in measuring pain was 0.96. The reported daily assessment or engagement rate ranged from 61.9% to 76.8%. All mHealth apps were designed for multimodal interventions. Participants generally had positive experiences using pain apps, rating them as enjoyable and user-friendly. In addition, 6 studies reported significant improvements in health outcomes, including enhancement in pain remission (severity and intensity), medication adherence, and a reduced frequency of breakthrough pain. The most frequently highlighted roles of mHealth apps included pain monitoring, tracking, reminders, education facilitation, and support coordination.
Conclusions: mHealth apps are effective and acceptable in supporting pain self-management. They offer a promising multi-model approach for patients to monitor, track, and manage their pain. These findings provide evidence-based insights for leveraging mHealth apps to support cancer pain self-management. More high-quality studies are needed to examine the effectiveness of digital technology-based interventions for cancer pain self-management and to identify the facilitators and barriers to their implementation in real-world practice.
Background: Personal sensing, leveraging data passively and near-continuously collected with wearables from patients in their ecological environment, is a promising paradigm to monitor mood disorders (MDs), a major determinant of the worldwide disease burden. However, collecting and annotating wearable data is resource intensive. Studies of this kind can thus typically afford to recruit only a few dozen patients. This constitutes one of the major obstacles to applying modern supervised machine learning techniques to MD detection.
Objective: In this paper, we overcame this data bottleneck and advanced the detection of acute MD episodes from wearables' data on the back of recent advances in self-supervised learning (SSL). This approach leverages unlabeled data to learn representations during pretraining, subsequently exploited for a supervised task.
Methods: We collected open access data sets recording with the Empatica E4 wristband spanning different, unrelated to MD monitoring, personal sensing tasks-from emotion recognition in Super Mario players to stress detection in undergraduates-and devised a preprocessing pipeline performing on-/off-body detection, sleep/wake detection, segmentation, and (optionally) feature extraction. With 161 E4-recorded subjects, we introduced E4SelfLearning, the largest-to-date open access collection, and its preprocessing pipeline. We developed a novel E4-tailored transformer (E4mer) architecture, serving as the blueprint for both SSL and fully supervised learning; we assessed whether and under which conditions self-supervised pretraining led to an improvement over fully supervised baselines (ie, the fully supervised E4mer and pre-deep learning algorithms) in detecting acute MD episodes from recording segments taken in 64 (n=32, 50%, acute, n=32, 50%, stable) patients.
Results: SSL significantly outperformed fully supervised pipelines using either our novel E4mer or extreme gradient boosting (XGBoost): n=3353 (81.23%) against n=3110 (75.35%; E4mer) and n=2973 (72.02%; XGBoost) correctly classified recording segments from a total of 4128 segments. SSL performance was strongly associated with the specific surrogate task used for pretraining, as well as with unlabeled data availability.
Conclusions: We showed that SSL, a paradigm where a model is pretrained on unlabeled data with no need for human annotations before deployment on the supervised target task of interest, helps overcome the annotation bottleneck; the choice of the pretraining surrogate task and the size of unlabeled data for pretraining are key determinants of SSL success. We introduced E4mer, which can be used for SSL, and shared the E4SelfLearning collection, along with its preprocessing pipeline, which can foster and expedite future research into SSL for personal sensing.
Background: Wearable activity trackers have become key players in mobile health practice as they offer various behavior change techniques (BCTs) to help improve physical activity (PA). Typically, multiple BCTs are implemented simultaneously in a device, making it difficult to identify which BCTs specifically improve PA.
Objective: We investigated the effects of BCTs implemented on a smartwatch, the Fitbit, to determine how each technique promoted PA.
Methods: This study was a single-blind, pilot randomized controlled trial, in which 70 adults (n=44, 63% women; mean age 40.5, SD 12.56 years; closed user group) were allocated to 1 of 3 BCT conditions: self-monitoring (feedback on participants' own steps), goal setting (providing daily step goals), and social comparison (displaying daily steps achieved by peers). Each intervention lasted for 4 weeks (fully automated), during which participants wore a Fitbit and responded to day-to-day questionnaires regarding motivation. At pre- and postintervention time points (in-person sessions), levels and readiness for PA as well as different aspects of motivation were assessed.
Results: Participants showed excellent adherence (mean valid-wear time of Fitbit=26.43/28 days, 94%), and no dropout was recorded. No significant changes were found in self-reported total PA (dz<0.28, P=.40 for the self-monitoring group, P=.58 for the goal setting group, and P=.19 for the social comparison group). Fitbit-assessed step count during the intervention period was slightly higher in the goal setting and social comparison groups than in the self-monitoring group, although the effects did not reach statistical significance (P=.052 and P=.06). However, more than half (27/46, 59%) of the participants in the precontemplation stage reported progress to a higher stage across the 3 conditions. Additionally, significant increases were detected for several aspects of motivation (ie, integrated and external regulation), and significant group differences were identified for the day-to-day changes in external regulation; that is, the self-monitoring group showed a significantly larger increase in the sense of pressure and tension (as part of external regulation) than the goal setting group (P=.04).
Conclusions: Fitbit-implemented BCTs promote readiness and motivation for PA, although their effects on PA levels are marginal. The BCT-specific effects were unclear, but preliminary evidence showed that self-monitoring alone may be perceived demanding. Combining self-monitoring with another BCT (or goal setting, at least) may be important for enhancing continuous engagement in PA.
Trial registration: Open Science Framework; https://osf.io/87qnb/?view_only=f7b72d48bb5044eca4b8ce729f6b403b.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：