Cross-Attention Enhanced Pyramid Multi-Scale Networks for Sensor-Based Human Activity Recognition.

Human Activity Recognition (HAR) has recently attracted widespread attention, with the effective application of this technology helping people in areas such as healthcare, smart homes, and gait analysis. Deep learning methods have shown remarkable performance in HAR. A pivotal challenge is the trade-off between recognition accuracy and computational efficiency, especially in resource-constrained mobile devices. This challenge necessitates the development of models that enhance feature representation capabilities without imposing additional computational burdens. Addressing this, we introduce a novel HAR model leveraging deep learning, ingeniously designed to navigate the accuracy-efficiency trade-off. The model comprises two innovative modules: 1) Pyramid Multi-scale Convolutional Network (PMCN), which is designed with a symmetric structure and is capable of obtaining a rich receptive field at a finer level through its multiscale representation capability; 2) Cross-Attention Mechanism, which establishes interrelationships among sensor dimensions, temporal dimensions, and channel dimensions, and effectively enhances useful information while suppressing irrelevant data. The proposed model is rigorously evaluated across four diverse datasets: UCI, WISDM, PAMAP2, and OPPORTUNITY. Additional ablation and comparative studies are conducted to comprehensively assess the performance of the model. Experimental results demonstrate that the proposed model achieves superior activity recognition accuracy while maintaining low computational overhead.

Diagnostic utility of prolonged ambulatory video-electroencephalography monitoring.

OBJECTIVES: Ambulatory video-electroencephalography (video-EEG) represents a low-cost, convenient and accessible alternative to inpatient video-EEG monitoring, however few studies have examined their diagnostic yield. In this large-scale retrospective study conducted in Australia, we evaluated the efficacy of prolonged ambulatory video-EEG recordings in capturing diagnostic events and resolving the referring question. METHODS: Sequential adult and paediatric ambulatory video-EEG reports from April 2020 to June 2021 were reviewed retrospectively. Data collection included patient demographics, clinical information, and details of events and EEG abnormalities. Clinical utility was assessed by examining i) time to first diagnostic event, and ii) ability to resolve the referring questions - seizure localisation, quantification, classification, and differentiation (differentiating seizures from non-epileptic events). RESULTS: Of the 600 reports analysed, 49 % captured at least one event, and 45 % captured interictal abnormalities (epileptiform or non-epileptiform). Seizures, probable psychogenic events (mostly non-convulsive), and other non-epileptic events occurred in 13 %, 23 % and 21 % of recordings respectively, with overlap. Unreported events were captured in 53 (9 %) recordings, and unreported seizures represented more than half of all seizures captured (51 %, 392/773). Nine percent of events were missing clinical, video or electrographic data. A diagnostic event occurred in 244 (41 %) recordings, of which 14 % were captured between the fifth and eighth day of recording. Reported event frequency ≥ 1/week was the only significant predictor of diagnostic event capture. In recordings with both seizures and psychogenic events, unrecognized seizures were frequent, and seizures may be missed if recording is terminated early. The referring question was resolved in 85 % of reports with at least one event, and 53 % of all reports. Specifically, this represented 46 % of reports (235/512) for differentiation of events, and 75 % of reports (27/36) for classification of seizures. CONCLUSION: Ambulatory video-EEG recordings are of high diagnostic value in capturing clinically relevant events and resolving the referring clinical questions.

Early discontinuation of ambulatory vEEG among individuals with intellectual disabilities: A retrospective chart review.

OBJECTIVE: This retrospective chart review aims to quantify the rate of patients with intellectual disability (ID) accessing an Australian ambulatory EEG service, and understand the clinical implications of discontinuing studies prematurely. METHODS: Electronic records of referrals, patient monitoring notes, and EEG reports were accessed retrospectively. Each referral was assessed to determine whether the patient had an ID. For each study where patients were discharged prematurely, the outcomes of their EEG report were assessed and compared between the ID and non-ID groups. Exploratory analysis was performed assessing the effects of age, the percentage of the requested monitoring undertaken, and outcome rates as a function of monitoring duration. RESULTS: There were significantly more patients in the ID group with early disconnection than the non-ID group (Chi squared test, p = 0.000). There was no significant difference in the rates of clinical outcomes between the ID and non-ID groups amongst patients who disconnected early. CONCLUSIONS: Although rates of early disconnection are higher in those with ID, study outcomes are largely similar between patients with and without ID in this retrospective analysis of an ambulatory EEG service. SIGNIFICANCE: Ambulatory EEG is a viable modality of EEG monitoring for patients with ID.

Validation of ActiGraph and Fitbit in the assessment of energy expenditure in Huntington's disease.

BACKGROUND: Consumer and research activity monitors have become popular because of their ability to quantify energy expenditure (EE) in free-living conditions. However, the accuracy of activity trackers in determining EE in people with Huntington's Disease (HD) is unknown. RESEARCH QUESTION: Can the ActiGraph wGT3X-B or the Fitbit Charge 4 accurately measure energy expenditure during physical activity, in people with HD compared to Indirect Calorimetry (IC) (Medisoft Ergo Card)? METHODS: We conducted a cross-sectional, observational study with fourteen participants with mild-moderate HD (mean age 55.7 ± 11.4 years). All participants wore an ActiGraph and Fitbit during an incremental test, running on a treadmill at 3.2 km/h and 5.2 km/h for three minutes at each speed. We analysed and compared the accuracy of EE estimates obtained by Fitbit and ActiGraph against the EE estimates obtained by a metabolic cart, using with Intra-class correlation (ICC), Bland-Altman analysis and correlation tests. RESULTS: A significant correlation and a moderate reliability was found between ActiGraph and IC for the incremental test (r = 0.667)(ICC=0.633). There was a significant correlation between Fitbit and IC during the incremental test (r = 0.701), but the reliability was poor at all tested speeds in the treadmill walk. Fitbit significantly overestimated EE, and ActiGraph underestimated EE compared to IC, but ActiGraph estimates were more accurate than Fitbit in all tests. SIGNIFICANCE: Compared to IC, Fitbit Charge 4 and ActiGraph wGT3X-BT have reduced accuracy in estimating EE at slower walking speeds. These findings highlight the need for population-specific algorithms and validation of activity trackers.

Ambulatory EEG-video.

Hospital based EEG recordings have been the norm to assist in the diagnosis and management of patients with unclassified events and known drug resistant epilepsy. Ambulatory EEG (AEEG) is a tool that comes to serve the needs for a portable testing that can be done at home, often with higher accessibility compared to an epilepsy monitoring unit and with lower cost. The current technology provides good quality EEG tracing and can be done with video when needed. In this review we discuss how AEEG should be performed and the preferred indications in which this test may be of utmost help. The advent of ultra-long ambulatory recording may be the future for selected patients as this technology evolves.

Diagnostic Yield of Ambulatory Reflux Monitoring Systems for Evaluation of Chronic Laryngeal Symptoms.

INTRODUCTION: Among patients with chronic laryngeal symptoms, ambulatory reflux monitoring off acid suppression is recommended to evaluate for laryngopharyngeal reflux (LPR). However, reflux monitoring systems are diverse in configuration and monitoring capabilities, which present a challenge in creating a diagnostic reference standard in these patients. This study aimed to compare diagnostic yield and performance between reflux monitoring systems in patients with chronic laryngeal symptoms. METHODS: This multicenter, international study of adult patients referred for evaluation of LPR over a 5-year period (March 2018-May 2023) assessed and compared diagnostic yield of pathologic gastroesophageal reflux (GER+) on ambulatory reflux monitoring off acid suppression. RESULTS: Of 813 patients, 296 (36%) underwent prolonged wireless pH, 532 (65%) underwent 24-hour pH-impedance monitoring, and 15 (2%) underwent both tests. Overall diagnostic yield for GER+ was 36% and greater for prolonged wireless pH compared with that for 24-hour pH-impedance monitoring (50% vs 27%; P < 0.01). Among 15 patients who underwent both prolonged wireless pH and 24-h pH-impedance monitoring, concordance between systems for GER+ was 40%. The most common source of discordance was strong evidence of GER+ across multiple days on prolonged wireless pH compared with no evidence of GER+ on pH-impedance. DISCUSSION: In this multicenter international study of patients with chronic laryngeal symptoms referred for LPR evaluation, diagnostic yield of ambulatory reflux monitoring off acid suppression was 36% and rose to 50% when using wireless pH monitoring. In patients referred for chronic laryngeal symptoms, 24-hour pH-impedance monitoring may risk a low negative predictive value in patients with unproven GER+ disease.

Cloud Native Remote Monitoring Data Ecosystem for Aging Population based on Commercial AAL Sensors.

In the recent years, Active Assisted Living (AAL) technologies used for autonomous tracking and activity recognition have started to play major roles in geriatric care. From fall detection to remotely monitoring behavioral patterns, vital functions and collection of air quality data, AAL has become pervasive in the modern era of independent living for the elderly section of the population. However, even with the current rate of progress, data access and data reliability has become a major hurdle especially when such data is intended to be used in new age modelling approaches such as those using machine learning. This paper presents a comprehensive data ecosystem comprising remote monitoring AAL sensors along with extensive focus on cloud native system architecture, secured and confidential access to data with easy data sharing. Results from a validation study illustrate the feasibility of using this system for remote healthcare surveillance. The proposed system shows great promise in multiple fields from various AAL studies to development of data driven policies by local governments in promoting healthy lifestyles for the elderly alongside a common data repository that can be beneficial to other research communities worldwide.Clinical Relevance- This study creates a cloud-based smart home data ecosystem, which can achieve the remote healthcare monitoring for aging population, enabling them to live more independently and decreasing hospital admission rates.

Cost-effectiveness of remote patient monitoring for First Nations peoples living with diabetes in regional Australia.

The aim of this study was to determine the cost-effectiveness of remote patient monitoring (RPM) with First Nations peoples living with diabetes. This study was set at the Goondir Health Service (GHS), an Aboriginal and Torres Strait Islander Community-Controlled Health in South-West Queensland. Electronic medical records and RPM data were provided by the GHS. Clinical effectiveness was determined by comparing mean HbA1c before and after enrolment in the RPM service. Our analysis found no statistically significant effect between the mean HbA1c before and after enrolment, so this analysis focused on net-benefit and return on investment for costs from the perspective of the GHS. The 6-month RPM service for 84 clients cost AUD $67,841 to cover RPM equipment, ongoing technology costs, and a dedicated Virtual Care Manager, equating to $808 per client. There were 199 additional client-clinician interactions in the period after enrolment resulting in an additional $4797 revenue for the GHS. Therefore, the program cost the GHS $63,044 to deliver, representing a return on investment of around 7 cents for every dollar they spent. Whilst the diabetes RPM service was equally effective as usual care and resulted in increased interactions with clients, the cost for the service was substantially more than the additional revenue generated from increased interactions. This evidence highlights the need for alternative funding models for RPM services and demonstrates the need to focus future research on long-term clinical effects and the extra-clinical benefits resulting from services of this type.

How often do healthy people cough?

Objective cough frequency has been reported in several respiratory conditions but the amount that healthy individuals cough daily is unclear. Seventy-nine healthy volunteers (38 males, median [IQR] age 41y [IQR 30-53]) completed 24-hour ambulatory cough monitoring (VitaloJAK™). The audio recording was filtered using a custom written algorithm to remove non-cough sounds and then all individual explosive cough sounds in the filtered file were tagged electronically by trained cough counters. Most coughing occurred during the day and cough numbers over 24 h were generally low (geometric mean of 4.6 coughs) but there was large variability; ranging from 0 to 136 coughs overall. Cough frequency was independent of participant characteristics apart from sex with males coughing significantly, 4-5 fold, more than females during the day and over 24 h (median [IQR] 16.1 [3.8-33.4] vs. 4.1 [1.0-15.0] total coughs; p = 0.015). This is the first report to describe cough frequency in a balanced group of healthy adults using an accurate cough monitoring system. The data reveal a further example of sexual dimorphism in cough, which warrants additional investigation.

PIPTO: Precise Inertial-Based Pipeline for Threshold-Based Fall Detection Using Three-Axis Accelerometers.

After traffic-related incidents, falls are the second cause of human death, presenting the highest percentage among the elderly. Aiming to address this problem, the research community has developed methods built upon different sensors, such as wearable, ambiance, or hybrid, and various techniques, such as those that are machine learning- and heuristic based. Concerning the models used in the former case, they classify the input data between fall and no fall, and specific data dimensions are required. Yet, when algorithms that adopt heuristic techniques, mainly using thresholds, are combined with the previous models, they reduce the computational cost. To this end, this article presents a pipeline for detecting falls through a threshold-based technique over the data provided by a three-axis accelerometer. This way, we propose a low-complexity system that can be adopted from any acceleration sensor that receives information at different frequencies. Moreover, the input lengths can differ, while we achieve to detect multiple falls in a time series of sum vector magnitudes, providing the specific time range of the fall. As evaluated on several datasets, our pipeline reaches high performance results at 90.40% and 91.56% sensitivity on MMsys and KFall, respectively, while the generated specificity is 93.96% and 85.90%. Lastly, aiming to facilitate the research community, our framework, entitled PIPTO (drawing inspiration from the Greek verb "πι´πτω", signifying "to fall"), is open sourced in Python and C.

Validity of activity wristbands for estimating daily physical activity in primary schoolchildren under free-living conditions: School-Fit study.

Introduction: The use of activity wristbands to monitor and promote schoolchildren's physical activity (PA) is increasingly widespread. However, their validity has not been sufficiently studied, especially among primary schoolchildren. Consequently, the main purpose was to examine the validity of the daily steps and moderate-to-vigorous PA (MVPA) scores estimated by the activity wristbands Fitbit Ace 2, Garmin Vivofit Jr 2, and the Xiaomi Mi Band 5 in primary schoolchildren under free-living conditions. Materials and methods: An initial sample of 67 schoolchildren (final sample = 62; 50% females), aged 9-12 years old (mean = 10.4 ± 1.0 years), participated in the present study. Each participant wore three activity wristbands (Fitbit Ace 2, Garmin Vivofit Jr 2, and Xiaomi Mi Band 5) on his/her non-dominant wrist and a research-grade accelerometer (ActiGraph wGT3X-BT) on his/her hip as the reference standard (number of steps and time in MVPA) during the waking time of one day. Results: Results showed that the validity of the daily step scores estimated by the Garmin Vivofit Jr 2 and Xiaomi Mi Band 5 were good and acceptable (e.g., MAPE = 9.6/11.3%, and lower 95% IC of ICC = 0.87/0.73), respectively, as well as correctly classified schoolchildren as meeting or not meeting the daily 10,000/12,000-step-based recommendations, obtaining excellent/good and good/acceptable results (e.g., Garmin Vivofit Jr 2, k = 0.75/0.62; Xiaomi Mi Band 5, k = 0.73/0.53), respectively. However, the Fitbit Ace 2 did not show an acceptable validity (e.g., daily steps: MAPE = 21.1%, and lower 95% IC of ICC = 0.00; step-based recommendations: k = 0.48/0.36). None of the three activity wristbands showed an adequate validity for estimating daily MVPA (e.g., MAPE = 36.6-90.3%, and lower 95% IC of ICC = 0.00-0.41) and the validity for the MVPA-based recommendation tended to be considerably lower (e.g., k = -0.03-0.54). Conclusions: The activity wristband Garmin Vivofit Jr 2 obtained the best validity for monitoring primary schoolchildren's daily steps, offering a feasible alternative to the research-grade accelerometers. Furthermore, this activity wristband could be used during PA promotion programs to provide accurate feedback to primary schoolchildren to ensure their accomplishment with the PA recommendations.

Signal Processing Circuits and Systems for Smart Sensing Applications.

The rising demand for reliable, real-time, low-maintenance, cost-efficient monitoring systems with a high accuracy is becoming increasingly more notable in everyday life [...].

Ambulatory video EEG extended to 10 days: A retrospective review of a large database of ictal events.

OBJECTIVE: This work aims to determine the ambulatory video electroencephalography monitoring (AVEM) duration and number of captured seizures required to resolve different clinical questions, using a retrospective review of ictal recordings. METHODS: Patients who underwent home-based AVEM had event data analyzed retrospectively. Studies were grouped by clinical indication: differential diagnosis, seizure type classification, or treatment assessment. The proportion of studies where the conclusion was changed after the first seizure was determined, as was the AVEM duration needed for at least 99% of studies to reach a diagnostic conclusion. RESULTS: The referring clinical question was not answered entirely by the first event in 29.6% (n = 227) of studies. Diagnostic and classification indications required a minimum of 7 days for at least 99% of studies to be answered, whilst treatment-assessment required at least 6 days. CONCLUSIONS: At least 7 days of monitoring, and potentially multiple events, are required to adequately answer these clinical questions in at least 99% of patients. The widely applied 72 h or single event recording cut-offs may be inadequate to adequately answer these three indications in a substantial proportion of patients. SIGNIFICANCE: Extended duration of monitoring and capturing multiple events should be considered when attempting to capture seizures on video-EEG.

Home monitoring for clinically suspected obstructive sleep apnea in pregnancy.

STUDY OBJECTIVES: To determine if a home sleep apnea test (HSAT) using a type III portable monitor (PM), Nox-T3 (Nox Medical, Inc., Reykjavik, Iceland), detects obstructive sleep apnea in pregnant women. METHODS: Ninety-two pregnant women (34.5 ± 4.3 years; gestational age 25.4 ± 8.9 weeks; body mass index 29.9 ± 4.7 kg/m2) with suspected obstructive sleep apnea underwent HSAT with the Nox-T3 PM followed by overnight polysomnography (PSG) and PM recording simultaneously in the laboratory within 1 week. PMs were scored automatically and manually using a 3% criteria and compared with PSGs scored by following guidelines. RESULTS: Apnea-hypopnea indexes were 8.56 ± 10.42, 8.19 ± 13.79, and 8.71 ± 14.19 events/h on HSAT, in-laboratory PM recording, and PSG (P = .955), respectively. Bland-Altman analysis of the apnea-hypopnea index on PSG vs HSAT showed a mean difference (95% confidence interval) of -0.15 (-1.83, 1.53); limits of agreement (± 2 SD) were -16.26 to 16.56 events/h. Based on a threshold apnea-hypopnea index ≥ 5 events/h, HSAT had 91% sensitivity, 85% specificity, 84% positive-predictive value, and 92% negative-predictive value compared with PSG. When comparing the simultaneous recordings, closer agreement was observed. Automated vs manual analysis of PM showed no significant difference. CONCLUSIONS: A type III PM had an acceptable failure rate and high diagnostic performance operating as a reasonable alternative for in-laboratory PSG in pregnant women. CITATION: Wang J, Zhang C, Xu L, et al. Home monitoring for clinically suspected obstructive sleep apnea in pregnancy. J Clin Sleep Med. 2023;19(11):1951-1960.

Robust Arm Impedocardiography Signal Quality Enhancement Using Recursive Signal Averaging and Multi-Stage Wavelet Denoising Methods for Long-Term Cardiac Contractility Monitoring Armbands.

Impedance cardiography (ICG) is a low-cost, non-invasive technique that enables the clinical assessment of haemodynamic parameters, such as cardiac output and stroke volume (SV). Conventional ICG recordings are taken from the patient's thorax. However, access to ICG vital signs from the upper-arm brachial artery (as an associated surrogate) can enable user-convenient wearable armband sensor devices to provide an attractive option for gathering ICG trend-based indicators of general health, which offers particular advantages in ambulatory long-term monitoring settings. This study considered the upper arm ICG and control Thorax-ICG recordings data from 15 healthy subject cases. A prefiltering stage included a third-order Savitzky-Golay finite impulse response (FIR) filter, which was applied to the raw ICG signals. Then, a multi-stage wavelet-based denoising strategy on a beat-by-beat (BbyB) basis, which was supported by a recursive signal-averaging optimal thresholding adaptation algorithm for Arm-ICG signals, was investigated for robust signal quality enhancement. The performance of the BbyB ICG denoising was evaluated for each case using a 700 ms frame centred on the heartbeat ICG pulse. This frame was extracted from a 600-beat ensemble signal-averaged ICG and was used as the noiseless signal reference vector (gold standard frame). Furthermore, in each subject case, enhanced Arm-ICG and Thorax-ICG above a threshold of correlation of 0.95 with the noiseless vector enabled the analysis of beat inclusion rate (BIR%), yielding an average of 80.9% for Arm-ICG and 100% for Thorax-ICG, and BbyB values of the ICG waveform feature metrics A, B, C and VET accuracy and precision, yielding respective error rates (ER%) of 0.83%, 11.1%, 3.99% and 5.2% for Arm-IG, and 0.41%, 3.82%, 1.66% and 1.25% for Thorax-ICG, respectively. Hence, the functional relationship between ICG metrics within and between the arm and thorax recording modes could be characterised and the linear regression (Arm-ICG vs. Thorax-ICG) trends could be analysed. Overall, it was found in this study that recursive averaging, set with a 36 ICG beats buffer size, was the best Arm-ICG BbyB denoising process, with an average of less than 3.3% in the Arm-ICG time metrics error rate. It was also found that the arm SV versus thorax SV had a linear regression coefficient of determination (R2) of 0.84.

[Design and Research of Wearable Fall Protection Device for the Elderly].

A protective device was designed that can be worn on the elderly, which consists of protective airbag, control box and protective mechanism. The combined acceleration, combined angular velocity and human posture angle are selected as the parameters to determine the fall, and the threshold algorithm and SVM algorithm are used to detect the fall. The protective mechanism is an inflatable device based on CO2 compressed air cylinder, and the equal-width cam structure is applied to its transmission part to improve the puncture efficiency of the compressed gas cylinder. A fall experiment was designed to obtain the combined acceleration and angular velocity eigenvalues of fall actions (forward fall, backward fall and lateral fall) and daily activities (sitting-standing, walking, jogging and walking up and down stairs), showing that the specificity and sensitivity of the protection module reached 92.1% and 84.4% respectively, which verified the feasibility of the fall protection device.