It is generally understood that wound areas appear smaller when calculated using 2D images, but the factors contributing to this discrepancy are not well-defined. With the rise of 3D photography, 3D segmentation, and 3D measurement, more accurate assessments have become possible. We developed an application called the Burn Evaluation Network (B.E.N.), which combines a deep learning model with LiDAR technology to perform both 2D and 3D measurements. In the first part of our study, we used burn wound templates to verify that the results of 3D segmentation closely matched the actual size of the burn wound and to examine the effect of limb curvature on the 3D/2D area ratio. Our findings revealed that smaller curvatures, indicative of flatter surfaces, were associated with lower 3D/2D area ratios, and larger curvatures corresponded to higher ratios. For instance, the back had the lowest average curvature (0.027 ± 0.004) and the smallest 3D/2D area ratio (1.005 ± 0.055). In the second part of our study, we applied our app to real patients, measuring burn areas in both 3D and 2D. Regions such as the head and neck (ratio: 1.641) and dorsal foot (ratio: 1.908) exhibited significantly higher 3D/2D area ratios. Additionally, images containing multiple burn wounds also showed a larger ratio (1.656) and greater variability in distribution. These findings suggest that 2D segmentation tends to significantly underestimate surface areas in highly curved regions or when measurements require summing multiple wound areas. We recommend using 3D measurements for wounds located on areas like the head, neck, and dorsal foot, as well as for cases involving multiple wounds or large areas, to improve measurement accuracy.

Background:The health impact of summer heat on older adults is a growing public concern, yet the physiological responses, particularly changes in resting heart rate (RHR), and the role of personal heat adaptation behaviors remain underexplored. Wearable devices offer an opportunity to objectively monitor physiological responses and evaluate the effectiveness of adaptation strategies in real-world settings. Objective:This study aimed to quantify the short-term association between summer temperatures and RHR in older adults and to examine how individual heat adaptation behaviors modify this relationship, with additional consideration of personal characteristics such as age, sex, BMI, and chronic disease status. Methods:We conducted a panel study among 83 community-dwelling older adults (≥65 y) in Taipei City during the summer of 2021 (May to September). Participants wore Garmin smartwatches to continuously monitor heart rate. Daily RHR was defined as the lowest 30-minute average heart rate. In September, heat adaptation behaviors were assessed via structured telephone interviews. Ambient temperature and relative humidity were obtained from a nearby monitoring station. Linear mixed-effect models were used to estimate temperature-RHR associations, and interaction terms were included to examine behavioral modifications. Subgroup analyses were conducted to explore effect modification by individual characteristics such as age, sex, BMI, and chronic disease status. Results:Each 1 °C increase in daily mean temperature over lag days 0-1 was associated with a 0.11 (95% CI 0.07-0.15; P<.001) beats/min increase in RHR. After mutual adjustment for behaviors, several heat adaptation strategies showed significant protective effects, including reducing physical activity (β=-.15, P=.001), drinking cold beverages (β=-.24, P<.001), increasing naps or sleep duration (β=-.28, P=.003), drinking additional water ≥500 mL (β=-.10, P=.02), using air conditioner (AC) before (β=-.15, P=.002) and during sleep (β=-.13, P=.007), and using electric fans during sleep (β=-.12, P=.01). Subgroup analyses revealed stronger effects for certain behaviors in vulnerable populations: reduced physical activity was particularly beneficial for those with higher BMI; AC use and cold beverage intake were more effective in people with diabetes; increased naps yielded the largest benefits in individuals with hypertension; and the use of AC or fans during sleep was especially protective for older adults and females. Conclusions:Summer heat is associated with elevated RHR in older adults, but this effect can be mitigated through targeted heat adaptation behaviors. Smartwatch monitoring provides a feasible and informative approach for capturing physiological changes, supporting the development of personalized heat-health recommendations for aging populations in a warming climate.

Falling can lead to bodily harm or even death for humans, so seeking immediate medical attention after a fall is crucial to minimize potential damage. However, it’s important to avoid unnecessary alarms; for instance, falling onto a bed versus falling onto the floor with the same posture and speed can have very different implications.We have developed a new fall detection system that utilizes LiDAR technology and the camera on a mobile device. LiDAR is used to gather 3D coordinates of the environment and assess the ground’s height based on the lowest point in the scene. Meanwhile, the camera and a pose estimation model capture the 2D coordinates of a person’s skeleton. By combining these data points, we can accurately identify a fall by analyzing the 3D coordinates of the person’s skeleton in relation to the ground’s height.This advanced system allows us to distinguish whether a person has fallen from an elevated surface to the ground, like a bed or sofa (which generally poses no harm), or on the ground (which may result in injury). The application can also determine if a person remains lying on a bed or sofa, considered a normal case with no alarm, versus lying on the ground, which is an abnormal case and warrants an alarm.Our study presents the first fall detection method that takes ground height into account, enabling real-time identification of falls using a modern mobile device and sending alerts accordingly with minimal false alarms.

Objectives: To predict the progression of knee osteoarthritis (OA), a deep convolutional neural network model was developed and applied to basic images and clinical data. Design: A vision transformer-based model was trained using 5565 knee radiographs as baseline images from the osteoarthritis initiative (OAI), including 578 testing images. Each knee had a corresponding Kellgren and Lawrence (KL) stage after 48 months of follow-up. Another 274 cases from the Far Eastern Memorial Hospital were used for external validation. The data included a combination of single/pairing images and full/essential clinical factors. Area under the receiver operating characteristics (AUROC), accuracy, sensitivity, specificity, odds ratio, and ability to discriminate surgical candidates were applied to evaluate model performance. Results: In cases with OA progression, the AUROC for identifying surgical candidates was 0.844, 0.804, 0.766, and 0.718 in the combination of a single image with essential factors, single image with full factors, pairing images with essential factors, and pairing images with full factors, respectively. In OAI testing using the simplest input, AUROC of identifying OA progression was 0.808, with 74.1% accuracy, 91.8% sensitivity, and 71% specificity. In external validation, AUROC of identifying OA progression was 0.709, with 71.2% accuracy, 72.2% sensitivity, and 70.3% specificity. Positive model prediction had an odds ratio of 23.87 (CI: 11.24~50.67) in OAI and 5.92 (CI: 3.50~10.03) in external validation. Conclusions: Our model provides reliable prediction results for knee OA cases with the advantages of simplicity and flexibility. The model performance was excellent in progression cases, potentially making early intervention in OA patients more efficient.

Background Severe community-acquired pneumonia was associated with high morbidity and mortality in children. However, species-level microbiome of lower airway was sparse, and we used shotgun metagenomic next-generation sequencing to explore microbial signatures. Methods We conducted a prospective cohort study to recruit children under 18 who required admission to an intensive care unit for community-acquired pneumonia between December 2019 and February 2022. Lower respiratory specimens were collected on admission for shotgun metagenomic sequencing. The children were divided into two groups. Critical cases were patients with respiratory failure requiring endotracheal ventilator support, and severe cases did not require intubation. Signatures of lower respiratory tract microbiome were compared between groups using an exact k-mer matching metagenomic analysis pipeline (Kraken 2) and a metagenome assembled genomes pipeline (MetaWRAP). Results Totally 66 children were enrolled, and 27 children were critical cases, and the rest were severe cases. There were significant differences in microbial community structure between different severity groups, and microbial abundance was negatively correlated with disease severity. The results showed that Haemophilus influenzae was more prominent in children who were critical, accompanied with increased expression of intracellular transport, secretion, and vesicle transport genes. Rothia mucilaginosa, Dolosigranulum pigrum, and Prevotella melaninogenica tended to be present in less severe community-acquired pneumonia group. Conclusion This study demonstrated that significantly different microbial community was associated with severity of community-acquired pneumonia requiring intensive care admission. Species-level shotgun metagenomic sequencing facilitates the exploration of potentially pathogenic or protective microbes and shed the light of probiotic development in lower respiratory tract.

BACKGROUND:Monitoring depressive symptoms in autistic individuals is challenging due to conditions in communication and emotional expression inherent to this population. For early detection of depressive symptoms in autistic adults, this study aims to leverage digital biomarkers from wearable devices and develop monitoring models using machine learning algorithms. METHODS:This prospective, observational study recruited 17 autistic adults (mean age 29.1 ± 8.2 years). Physiological biomarkers, including activity level, heart rate, and sleep duration, were continuously collected via smartwatches. Depressive symptoms were self-rated weekly using the Beck Depression Inventory (BDI). Machine learning models, including Extreme Gradient Boosting (XGBoost), were applied to analyze the longitudinal data. The models were assessed using the area under the receiver operating characteristic curve (AUROC), sensitivity, specificity, accuracy, precision, and F1 score. Feature importance analysis was conducted to identify key digital biomarkers. RESULTS:The XGBoost model demonstrated superior predictive efficacy, identifying depressive states with an 84 % accuracy and an area under the receiver operating curve (AUROC) of 0.91. Lower activity levels, decreased sleep duration, and reduced average heart rate emerged as potential predictors for depressive states. CONCLUSION:Digital phenotypes derived from wearable devices may facilitate the detection of depressive symptoms in autistic adults, offering potential benefits for clinical assessments and emotion self-care.

ObjectivesThis study aimed to predict mortality in children with pneumonia who were admitted to the intensive care unit (ICU) to aid decision-making.Study DesignRetrospective cohort study conducted at a single tertiary hospital.PatientsThis study included children who were admitted to the pediatric ICU at the National Taiwan University Hospital between 2010 and 2019 due to pneumonia.MethodologyTwo prediction models were developed using tree-structured machine learning algorithms. The primary outcomes were ICU mortality and 24-h ICU mortality. A total of 33 features, including demographics, underlying diseases, vital signs, and laboratory data, were collected from the electronic health records. The machine learning models were constructed using the development data set, and performance matrices were computed using the holdout test data set.ResultsA total of 1231 ICU admissions of children with pneumonia were included in the final cohort. The area under the receiver operating characteristic curves (AUROCs) of the ICU mortality model and 24-h ICU mortality models was 0.80 (95% confidence interval [CI], 0.69-0.91) and 0.92 (95% CI, 0.86-0.92), respectively. Based on feature importance, the model developed in this study tended to predict increased mortality for the subsequent 24 h if a reduction in the blood pressure, peripheral capillary oxygen saturation (SpO2), or higher partial pressure of carbon dioxide (PCO2) were observed.ConclusionsThis study demonstrated that the machine learning models for predicting ICU mortality and 24-h ICU mortality in children with pneumonia have the potential to support decision-making, especially in resource-limited settings.

The measurement of height is a crucial indicator for assessing health and making informed decisions in various fields. Using a depth camera for height measurement has become popular due to its convenience and efficiency, as it does not require physical contact. However, this method has limitations, such as needing an unobstructed view of the body. To address this, we propose a height measurement mechanism using Light Detection and Ranging (LiDAR) in combination with a camera. By using LiDAR, we can obtain three-dimensional (3D) coordinates of the real-world and determine the lowest point on the ground. Additionally, we use a camera and face detection method to obtain two-dimensional (2D) coordinates of a human face. By combining these data, we can estimate a person’s height, even if their body is obstructed, without requiring a reference length, specific camera position, or calibration. This method allows for the real-time measurement of multiple individuals using a modern mobile device with few limitations, making it suitable for use in crowded environments.

BACKGROUND:This study aimed to evaluate the feasibility of using mHealth devices for monitoring postoperative ambulation among patients with colorectal cancer undergoing minimally invasive surgery (MIS). METHODS:Patients with colorectal cancer undergoing MIS were prospectively recruited to wear mHealth devices for recording postoperative ambulation between October 2018 and January 2021. The primary outcome was the compliance by evaluating the weekly submission rate of step counts. The secondary outcome was the association of weekly step counts and postoperative length of stay. RESULTS:Of 107 eligible patients, 53 patients wore mHealth devices, whereas 54 patients did not. The average weekly submission rate was 72.6% for the first month after surgery. The total step counts <4000 or >10 000 in the postoperative week one were negatively associated with postoperative length of stay (β = -2.874, p = 0.038). CONCLUSIONS:mHealth devices provide an objective assessment of postoperative ambulation among patients with colorectal cancer undergoing MIS. CLINICAL TRIAL REGISTRATION:NCT03277235.

Background: Sleep and physical activity suggestions for panic disorder (PD) are critical but less surveyed. This two-year prospective cohort study aims to predict panic attacks (PA), state anxiety (SA), trait anxiety (TA) and panic disorder severity (PDS) in the upcoming week. Methods: We enrolled 114 PD patients from one general hospital. Data were collected using the DSM-5, the MINI, clinical app questionnaires (BDI, BAI, PDSS-SR, STAI) and wearable devices recording daily sleep, physical activity and heart rate from 16 June 2020 to 10 June 2022. Our teams applied RNN, LSTM, GRU deep learning and SHAP explainable methods to analyse the data. Results: The 7-day prediction accuracies for PA, SA, TA, and PDS were 92.8 %, 83.6 %, 87.2 %, and 75.6 % from the LSTM model. Using the SHAP explainable model, higher initial BDI or BAI score and comorbidities with depressive disorder, generalized anxiety disorder or agoraphobia predict a higher chance of PA. However, PA decreased under the following conditions: daily average heart rate, 72-87 bpm; maximum heart rate, 100-145 bpm; resting heart rate, 55-60 bpm; daily climbing of more than nine floors; total sleep duration between 6 h 23 min and 10 h 50 min; deep sleep, >50 min; and awake duration, <53 min. Limitations: Moderate sample size and self-report questionnaires were the limitations. Conclusions: Deep learning predicts recurrent PA and various anxiety domains with 75.6-92.8 % accuracy. Recurrent PA decreases under adequate daily sleep and physical activity.

Although neuropsychiatric manifestations are common in survivors of coronavirus disease 2019 (COVID-19), the pathophysiology is not yet elucidated. Here we describe the case of a geriatric inpatient who developed post-COVID depression with psychomotor retardation, anxiety, hopelessness, executive function problems, and suicidal ideations. The language problems and cognitive impairments coemerged with the motor problems. We propose a mechanism associated with problems in energy prediction and regulation in which the coronavirus infection, which causes neuroinflammation and viral activity in the nervous system, interferes with the reward pathway and sensory prediction process. Sigma-1 receptor agonists such as sertraline may regulate energy expenditure and, thus, be beneficial to the process. The treatment improvements in our patient included those in the autonomic nervous system, activity, and circadian rhythm.

Background: Epidemiological evidence regarding the association between air pollution and resting heart rate (RHR), a predictor of cardiovascular disease and mortality, is limited and inconsistent. Objectives: We used wearable devices and time-series analysis to assess the exposure-response relationship over an extended lag period. Methods: Ninety-seven elderly individuals (>65 years) from the Taipei Basin participated from May to November 2020 and wore Garmin (R) smartwatches continuously until the end of 2021 for heart rate monitoring. RHR was defined as the daily average of the lowest 30-min heart rate. Air pollution exposure data, covering lag periods from 0 to 60 days, were obtained from nearby monitoring stations. We used distributed lag non-linear models and linear mixed-effect models to assess cumulative effects of air pollution. Principal component analysis was utilized to explore underlying patterns in air pollution exposure, and subgroup analyses with interaction terms were conducted to explore the modification effects of individual factors. Results: After adjusting for co-pollutants in the models, an interquartile range increase of 0.18 ppm in carbon monoxide (CO) was consistently associated with increased RHR across lag periods of 0-1 day (0.31, 95 % confidence interval [CI]: 0.24-0.38), 0-7 days (0.68, 95 % CI: 0.57-0.79), and 0-50 days (1.02, 95 % CI: 0.82-1.21). Principal component analysis identified two factors, one primarily influenced by CO and nitrogen dioxide (NO2), indicative of traffic sources. Increases in the varimax-rotated traffic-related score were correlated with higher RHR over 0-1 day (0.36, 95 % CI: 0.25-0.47), 0-7 days (0.62, 95 % CI: 0.46-0.77), and 0-50 days (1.27, 95 % CI: 0.87-1.67) lag periods. Over a 0-7 day lag, RHR responses to traffic pollution were intensified by higher temperatures (beta = 0.80 vs. 0.29; interaction p-value [P_int] = 0.011). Males (beta = 0.66 vs. 0.60; P_int < 0.0001), hypertensive individuals (beta = 0.85 vs. 0.45; P_int = 0.028), diabetics (beta = 0.96 vs. 0.52; P_int = 0.042), and those with lower physical activity (beta = 0.70 vs. 0.54; P_int < 0.0001) also exhibited stronger responses. Over a 0-50 day lag, males (beta = 0.99 vs. 0.96; P_int < 0.0001), diabetics (beta = 1.66 vs. 0.69; P_int < 0.0001), individuals with lower physical activity (beta = 1.49 vs. 0.47; P_int = 0.0006), and those with fewer steps on lag day 1 (beta = 1.17 vs. 0.71; P_int = 0.029) showed amplified responses. Conclusions: Prolonged exposure to traffic-related air pollution results in cumulative cardiovascular risks, persisting for up to 50 days. These effects are more pronounced on warmer days and in individuals with chronic conditions or inactive lifestyles.

Purpose This study seeks to build a normative database for the vessel density of the superficial retina (SVD) and evaluate how changes and trends in the retinal microvasculature may be influenced by age and axial length (AL) in non-glaucomatous eyes, as measured with optical coherence tomography angiography (OCTA). Methods We included 500 eyes of 290 healthy subjects visiting a county hospital. Each participant underwent comprehensive ophthalmological examinations and OCTA to measure the SVD and thickness of the macular and peripapillary areas. To analyze correlations between SVD and age or AL, multivariable linear regression models with generalized estimating equations were applied. Results Age was negatively correlated with the SVD of the superior, central, and inferior macular areas and the superior peripapillary area, with a decrease rate of 1.06%, 1.36%, 0.84%, and 0.66% per decade, respectively. However, inferior peripapillary SVD showed no significant correlation with age. AL was negatively correlated with the SVD of the inferior macular area and the superior and inferior peripapillary areas, with coefficients of −0.522%/mm, −0.733%/mm, and −0.664%/mm, respectively. AL was also negatively correlated with the thickness of the retinal nerve fiber layer and inferior ganglion cell complex ( p = 0.004). Conclusion Age and AL were the two main factors affecting changes in SVD. Furthermore, AL, a relative term to represent the degree of myopia, had a greater effect than age and showed a more significant effect on thickness than on SVD. This relationship has important implications because myopia is a significant issue in modern cities.

Abstract Background We aimed to develop and validate models for predicting intensive care unit (ICU) mortality of critically ill adult patients as early as upon ICU admission. Methods Combined data of 79,657 admissions from two teaching hospitals’ ICU databases were used to train and validate the machine learning models to predict ICU mortality upon ICU admission and at 24 h after ICU admission by using logistic regression, gradient boosted trees (GBT), and deep learning algorithms. Results In the testing dataset for the admission models, the ICU mortality rate was 7%, and 38.4% of patients were discharged alive or dead within 1 day of ICU admission. The area under the receiver operating characteristic curve (0.856, 95% CI 0.845–0.867) and area under the precision-recall curve (0.331, 95% CI 0.323–0.339) were the highest for the admission GBT model. The ICU mortality rate was 17.4% in the 24-hour testing dataset, and the performance was the highest for the 24-hour GBT model. Conclusion The ADM models can provide crucial information on ICU mortality as early as upon ICU admission. 24 H models can be used to improve the prediction of ICU mortality for patients discharged more than 1 day after ICU admission.

Background: COVID-19 is notorious for its neuropsychiatric sequelae. Some patients present with anosmia and cognitive and attention deficits, also known as “brain-fog”. In COVID-19 survivors, psychiatric manifestations such as depression, fatigue, anxiety, and neurocognitive impairments, also known as the “post-acute COVID-19 syndrome” or “long COVID”, were reported in 35% of the patients after the infection. New-onset anxiety, depression, psychosis, seizures, and even suicidal behaviors are also reported. Inflammation was proposed to be the pathophysiology. Case Report: The 68 year-old married female who lived with her husband did not have psychiatric history before this year. She had covid-19 infection 3 months prior to her psychiatric hospitalization. Although she had recovered, she started to have depressed mood, disrupted sleep and anhedonia since then. She became easily anxious and could hardly feed her cats as usual. She was not able to make decisions. She no longer did house chores that she had been doing. Poor appetite led to a loss of body weight from 46 to 40kgs. She had blunted emotional response, and could not cry after her cat died. She went to clinic for help, BZD and antidepressants were given, but the condition persisted. One week prior to hospitalization, her speech became weak, murmuring and could not finish a long sentence. The physiological change can be observed using wearable device. After 3 weeks of treatment, her heat rate, activity and sleep improved. The progress was also recorded via wearable devices and clinical scales. Discussion: In our patient, the depression was associated with circadian rhythm disruption, suicidal ideations and psychomotor slowing. This may reveal some special features of post-COVID depression. In previous studies, antidepressants have proved to be effective in treating post-COVID-19 depression. Sigma receptors are one of the possible mechanisms. Besides, agomelatine, with its melatonergic activity of regulating circadian rhythm, is proposed to have the potential to inhibit the virus from entering CNS. The activity pattern and circadian rhythm change will be discussed.

BackgroundIndividuals with panic disorder frequently face ongoing symptoms, suboptimal treatment adherence, and increased relapse rates. Although mobile health interventions have shown promise in improving treatment outcomes for numerous mental health conditions, their effectiveness, specifically for panic disorder, has yet to be determined.ObjectiveThis study investigates the effects of a mobile-aided case management program on symptom reduction and quality of care among individuals with panic disorder.MethodsThis 3-year cohort study enrolled 138 participants diagnosed with panic disorder. One hundred and eight participants joined the mobile-aided case management group and 30 in the treatment-as-usual group. Data were collected at baseline, 3-month, 6-month, and 12-month treatment checkpoints using self-report questionnaires, in-depth interviews, direct observation, and medical record analysis.ResultsDuring the maintenance treatment phase, the mobile-assisted case management group decreased both panic severity (p = 0.008) and state anxiety (p = 0.016) more than the control group at 6 months. Participants who underwent case management experienced enhanced control over panic symptoms, heightened self-awareness, and elevated interpersonal support.ConclusionThe mobile-aided case management is beneficial in managing panic disorder, especially maintenance treatment.

Pressure injuries are a common problem resulting in poor prognosis, long-term hospitalization, and increased medical costs in an aging society. This study developed a method to do automatic segmentation and area measurement of pressure injuries using deep learning models and a light detection and ranging (LiDAR) camera. We selected the finest photos of patients with pressure injuries, 528 in total, at National Taiwan University Hospital from 2016 to 2020. The margins of the pressure injuries were labeled by three board-certified plastic surgeons. The labeled photos were trained by Mask R-CNN and U-Net for segmentation. After the segmentation model was constructed, we made an automatic wound area measurement via a LiDAR camera. We conducted a prospective clinical study to test the accuracy of this system. For automatic wound segmentation, the performance of the U-Net (Dice coefficient (DC): 0.8448) was better than Mask R-CNN (DC: 0.5006) in the external validation. In the prospective clinical study, we incorporated the U-Net in our automatic wound area measurement system and got 26.2% mean relative error compared with the traditional manual method. Our segmentation model, U-Net, and area measurement system achieved acceptable accuracy, making them applicable in clinical circumstances.

Background: Acute respiratory infections (ARIs) are common in children. We devel-oped machine learning models to predict pediatric ARI pathogens at admission.Methods: We included hospitalized children with respiratory infections between 2010 and 2018. Clinical features were collected within 24 h of admission to construct models. The outcome of interest was the prediction of 6 common respiratory pathogens, including adeno-virus, influenza virus types A and B, parainfluenza virus (PIV), respiratory syncytial virus (RSV), and Mycoplasma pneumoniae (MP). Model performance was estimated using area under the receiver operating characteristic curve (AUROC). Feature importance was measured using Shapley Additive exPlanation (SHAP) values.Results: A total of 12,694 admissions were included. Models trained with 9 features (age, event pattern, fever, C-reactive protein, white blood cell count, platelet count, lymphocyte ratio, peak temperature, peak heart rate) achieved the best performance (AUROC: MP 0.87, 95% CI 0.83-0.90; RSV 0.84, 95% CI 0.82-0.86; adenovirus 0.81, 95% CI 0.77-0.84 ; influenza A 0.77, 95% CI 0.73-0.80; influenza B 0.70, 95% CI 0.65-0.75; PIV 0.73, 95% CI 0.69-0.77). Age was the most important feature to predict MP, RSV , PIV infections. Event patterns were useful for influenza virus prediction , C-reactive protein had the highest SHAP value for adenovirus infections.Conclusion: We demonstrate how artificial intelligence can assist clinicians identify potential pathogens associated with pediatric ARIs upon admission. Our models provide explainable re-sults that could help optimize the use of diagnostic testing. Integrating our models into clinical workflows may lead to improved patient outcomes and reduce unnecessary medical costs. Copyright 2023, Taiwan Society of Microbiology. Published by Elsevier Taiwan LLC. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Purpose The 6-minute walk test (6MWT) is often used to evaluate chronic obstructive pulmonary disease (COPD) patients’ functional capacity, with 6-minute walk distance (6MWD) and related measures being linked to mortality and hospitalizations. This study investigates the prognostic value of pace variability, a significant indicator in sports medicine, during the 6MWT for COPD patients. Patients and Methods We retrospectively screened consecutive COPD patients who had been prospectively enrolled in a pay-for-performance program from January 2019 to May 2020 to determine their eligibility. Patient characteristics, including demographics, exacerbation history, and 6MWT data, were analyzed to investigate their potential associations with prognosis. The primary outcome was a composite of adverse events, including overall mortality or hospitalizations due to exacerbations during a 1-year follow-up period. To analyze the 6MWT data, we divided it into three 2-minute epochs and calculated the average walk speed for each epoch. We defined pace variability as the difference between the maximum and minimum average speed in a single 2-minute epoch, divided by the average speed for the entire 6-minute walk test. Results A total of 163 patients with COPD were included in the study, and 19 of them (12%) experienced the composite adverse outcome. Multivariable logistic regression analyses revealed that two predictors were independently associated with the composite outcome: % predicted 6MWD <72 (adjusted odds ratio [aOR] 7.080; 95% confidence interval [CI] 1.481–33.847) and pace variability ≥0.39 (aOR 9.444; 95% CI 2.689–33.170). Patients with either of these adverse prognostic features had significantly worse composite outcome-free survival, with both log-rank P values less than 0.005. Notably, COPD patients with both adverse features experienced an especially poor outcome after 1 year. Conclusion Patients with COPD who exhibited greater pace variability during the 6MWT had a significantly higher risk of overall mortality and COPD-related hospitalizations, indicating a worse prognosis.

Introduction: Healthcare professionals frequently experience work-related fatigue, which may jeopardize their health and put patient safety at risk. In this study, we applied a machine learning (ML) approach based on data collected from a smartwatch to construct prediction models of work-related fatigue for emergency clinicians. Methods: We conducted this prospective study at the emergency department (ED) of a tertiary teaching hospital from March 10-June 20, 2021, where we recruited physicians, nurses, and nurse practitioners. All participants wore a commercially available smartwatch capable of measuring various physiological data during the experiment. Participants completed the Multidimensional Fatigue Inventory (MFI) web form before and after each of their work shifts. Wecalculated and labeled the before-and-after-shift score differences between each pair of scores. Using several tree-based algorithms, we constructed the prediction models based on features collected from the smartwatch. Records were split into training/validation and testing sets at a 70:30 ratio, and we evaluated the performances using the area under the curve (AUC) measure of receiver operating characteristic on the test set. Results: In total, 110 participants were included in this study, contributing to a set of 1,542 effective records. Of these records, 85 (5.5%) were labeled as having work-related fatigue when setting the MFI difference between two standard deviations as the threshold. The mean age of the participants was 29.6. Most of the records were collected from nurses (87.7%) and females (77.5%). We selected a union of 31 features to construct the models. For total participants, CatBoost classifier achieved the best performances of AUC (0.838, 95% confidence interval [CI] 0.742-0.918) to identify work-related fatigue. By focusing on a subgroup of nurses <35 years in age, XGBoost classifier obtained excellent performance of AUC (0.928, 95% CI 0.839-0.991) on the test set. Conclusion: By using features derived from a smartwatch, we successfully built ML models capable of classifying the risk of work-related fatigue in the ED. By collecting more data to optimize the models, it should be possible to use smartwatch-based ML models in the future to predict work-related fatigue and adopt preventive measures for emergency clinicians.