Coreference Resolution Based on High-Dimensional Multi-Scale Information.

Coreference resolution is a key task in Natural Language Processing. It is difficult to evaluate the similarity of long-span texts, which makes text-level encoding somewhat challenging. This paper first compares the impact of commonly used methods to improve the global information collection ability of the model on the BERT encoding performance. Based on this, a multi-scale context information module is designed to improve the applicability of the BERT encoding model under different text spans. In addition, improving linear separability through dimension expansion. Finally, cross-entropy loss is used as the loss function. After adding BERT and span BERT to the module designed in this article, F1 increased by 0.5% and 0.2%, respectively.

Modeling of the Rating of Perceived Exertion Based on Heart Rate Using Machine Learning Methods.

Rating of perceived exertion (RPE) can serve as a more convenient and economical alternative to heart rate (HR) for exercise intensity control. This study aims to explore the influence of factors, such as indicators of demographic, anthropometric, body composition, cardiovascular function and basic exercise ability on the relationship between HR and RPE, and to develop the model predicting RPE from HR. 48 healthy participants were recruited to perform an incrementally 6-stage pedaling test. HR and RPE were collected during each stage. The influencing factors were identified with the forward selection method to train Gaussian Process regression (GPR), support vector machine (SVM) and linear regression models. Metrics of R2, adjusted R2 and RMSE were calculated to evaluate the performance of the models. The GPR model outperformed the SVM and linear regression models, and achieved an R2 of 0.95, adjusted R2 of 0.89 and RMSE of 0.52. Indicators of age, resting heart rate (RHR), Central arterial pressure (CAP), body fat rate (BFR) and body mass index (BMI) were identified as factors that best predicted the relationship between RPE and HR. It is possible to use GPR model to estimate RPE from HR accurately, after adjusting for age, RHR, CAP, BFR and BMI.

Impact of Age, Gender, and Body Composition on Bone Quality in an Adult Population From the Middle Areas of China.

Identifying modifiable factors that influence bone status during adulthood to maximize bone quality is a potential primary strategy in the prevention of osteoporosis in later life. We investigated the impact of body height, body weight, body mass index, and body composition on calcaneal bone characteristics as measured with quantitative ultrasound in 441 Chinese adults (238 women) aged 20-55 yr from the middle areas of China. Body composition, including fat-free mass (FFM), muscle mass, and fat mass were obtained by bioelectrical impedance analysis. Bivariate correlation analysis demonstrated a significant negative correlation between age and broadband ultrasound attenuation (BUA), speed of sound (SOS), and stiffness index (SI) both in men (r = -0.177, p < 0.05; r = -0.499, p < 0.001; r = -0.530, p < 0.001, respectively) and women (r = -0.344, p < 0.001; r = -0.336, p < 0.001; r = -0.369, p < 0.001, respectively). Body height, body weight, FFM, and muscle mass had positive correlations with BUA, SOS, and SI in both genders, with FFM having the strongest correlation with BUA, SOS, and SI in men (r = 0.351, p < 0.001; r = 0.391, p < 0.001; r = 0.406, p < 0.001, respectively) and women (r = 0.331, p < 0.001; r = 0.288, p < 0.001; r = 0.324, p < 0.001, respectively). Fat mass had a positive correlation with BUA (r = 0.331, p < 0.001), SOS (r = 0.288, p < 0.001), and SI (r = 0.324, p < 0.001) in women, which was not found in men. Multivariate regression analysis revealed that, in both genders, FFM was a positive predictor for all 3 quantitative ultrasound variables.

Detection of mild cognitive impairment based on attention mechanism and parallel dilated convolution.

Mild cognitive impairment (MCI) is a precursor to neurodegenerative diseases such as Alzheimer's disease, and an early diagnosis and intervention can delay its progression. However, the brain MRI images of MCI patients have small changes and blurry shapes. At the same time, MRI contains a large amount of redundant information, which leads to the poor performance of current MCI detection methods based on deep learning. This article proposes an MCI detection method that integrates the attention mechanism and parallel dilated convolution. By introducing an attention mechanism, it highlights the relevant information of the lesion area in the image, suppresses irrelevant areas, eliminates redundant information in MRI images, and improves the ability to mine detailed information. Parallel dilated convolution is used to obtain a larger receptive field without downsampling, thereby enhancing the ability to acquire contextual information and improving the accuracy of small target classification while maintaining detailed information on large-scale feature maps. Experimental results on the public dataset ADNI show that the detection accuracy of the method on MCI reaches 81.63%, which is approximately 6.8% higher than the basic model. The method is expected to be used in clinical practice in the future to provide earlier intervention and treatment for MCI patients, thereby improving their quality of life.

Development of a Health Behavioral Digital Intervention for Patients With Hypertension Based on an Intelligent Health Promotion System and WeChat: Randomized Controlled Trial.

BACKGROUND: The effectiveness of timely medication, physical activity (PA), a healthy diet, and blood pressure (BP) monitoring for promoting health outcomes and behavioral changes among patients with hypertension is supported by a substantial amount of literature, with "adherence" playing a pivotal role. Nevertheless, there is a lack of consistent evidence regarding whether digital interventions can improve adherence to healthy behaviors among individuals with hypertension. OBJECTIVE: The aim was to develop a health behavioral digital intervention for hypertensive patients (HBDIHP) based on an intelligent health promotion system and WeChat following the behavior change wheel (BCW) theory and digital micro-intervention care (DMIC) model and assess its efficacy in controlling BP and improving healthy behavior adherence. METHODS: A 2-arm, randomized trial design was used. We randomly assigned 68 individuals aged >60 years with hypertension in a 1:1 ratio to either the control or experimental group. The digital intervention was established through the following steps: (1) developing digital health education materials focused on adherence to exercise prescriptions, Dietary Approaches to Stop Hypertension (DASH), prescribed medication, and monitoring of BP; (2) using the BCW theory to select behavior change techniques; (3) constructing the intervention's logic following the guidelines of the DMIC model; (4) creating an intervention manual including the aforementioned elements. Prior to the experiment, participants underwent physical examinations at the community health service center's intelligent health cabin and received intelligent personalized health recommendations. The experimental group underwent a 12-week behavior intervention via WeChat, while the control group received routine health education and a self-management manual. The primary outcomes included BP and adherence indicators. Data analysis was performed using SPSS, with independent sample t tests, chi-square tests, paired t tests, and McNemar tests. A P value <.05 was considered statistically significant. RESULTS: The final analysis included 54 participants with a mean age of 67.24 (SD 4.19) years (n=23 experimental group, n=31 control group). The experimental group had improvements in systolic BP (-7.36 mm Hg, P=.002), exercise time (856.35 metabolic equivalent [MET]-min/week, P<.001), medication adherence (0.56, P=.001), BP monitoring frequency (P=.02), and learning performance (3.23, P<.001). Both groups experienced weight reduction (experimental: 1.2 kg, P=.002; control: 1.11 kg, P=.009) after the intervention. The diet types and quantities for both groups (P<.001) as well as the subendocardial viability ratio (0.16, P=.01) showed significant improvement. However, there were no statistically significant changes in other health outcomes. CONCLUSIONS: The observations suggest our program may have enhanced specific health outcomes and adherence to health behaviors in older adults with hypertension. However, a longer-term, larger-scale trial is necessary to validate the effectiveness. TRIAL REGISTRATION: Chinese Clinical Trial Registry ChiCTR2200062643; https://www.chictr.org.cn/showprojEN.html?proj=172782. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR2-10.2196/46883.

Evaluating a WeChat-Based Health Behavioral Digital Intervention for Patients With Hypertension: Protocol for a Randomized Controlled Trial.

BACKGROUND: Hypertension is the most prevalent chronic condition and a significant risk factor for cardiovascular and kidney diseases. The efficacy of health behavioral interventions in blood pressure (BP) control has been demonstrated by a large and expanding body of literature, with "adherence" playing a crucial role. WeChat is the most common social communication mobile app in China, and it has been shown to be an acceptable delivery platform for delivering health interventions. The WeChat-based health behavioral digital intervention program (WHBDIP) showed high feasibility and efficacy. However, the results regarding BP improvement between the WHBDIP and control groups were inconsistent. OBJECTIVE: The objective of this study is to develop a WHBDIP and assess its efficacy in controlling BP and improving adherence among patients with hypertension. METHODS: A 2-arm, parallel-group, and randomized trial design was used. Patients older than 60 years and with hypertension were randomly assigned to either the control group or the experimental group, which received a 12-week intervention. The program, primarily developed based on the Behavior Change Wheel (BCW) theory, offers health education on exercise, diet, BP monitoring, and medicine adherence (MA). It also includes other behavior interventions guided by an intervention manual, incorporating behavior change techniques (BCTs). The primary outcomes encompass BP and adherence indicators, while the secondary outcomes encompass cardiovascular function indicators, body composition indicators, learning performance, satisfaction, and acceptability. The exercise and blood pressure monitoring adherence (BPMA) indicators for the WHBDIP group were assessed weekly via WeChat during the initial 3 months, while other outcome data for both groups will be collected at the baseline assessment phase, 3 months after the intervention, and 1 year after the program. RESULTS: The trial will assess the efficacy of WHBDIP for patients with hypertension (N=68). The WHBDIP seeks to enhance participants' knowledge of healthy behaviors and assist patients in developing positive health behaviors to improve their health outcomes. Patient recruitment for individuals with hypertension commenced on September 5, 2022, and concluded on September 19, 2022. The 3-month intervention and phased data collection were finalized in January 2023. Data analysis will commence in August 2023, and the final 1-year health outcome results will be collected in September 2023. CONCLUSIONS: A successful WHBDIP will establish the management mode as a feasible approach for hypertension management in the community. Additionally, it will pave the way for the development of related mobile health programs. TRIAL REGISTRATION: Chinese Clinical Trial Registry ChiCTR2200062643; https://tinyurl.com/mwyv67wk. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/46883.

A weighted cranial diffusion-weighted imaging scale for Wilson's disease.

Objectives: Cranial magnetic resonance imaging (MRI) could be a crucial tool for the assessment for neurological symptoms in patients with Wilson's disease (WD). Diffusion-weighted imaging (DWI) hyperintensity reflects the acute brain injuries, which mainly occur in specific brain regions. Therefore, this study aimed to develop a weighted cranial DWI scale for patients with WD, with special focus on specific brain regions. Materials and methods: In total, 123 patients with WD were enrolled, 118 of whom underwent 1.5 T-MRI on admission. The imaging score was calculated as described previously and depended on the following sequences: one point was acquired when abnormal intensity occurred in the T1, T2, and fluid-attenuation inversion recovery sequences, and two points were acquired when DWI hyperintensity were found. Consensus weighting was conducted based on the symptoms and response to treatment. Results: Intra-rater agreement were good (r = 0.855 [0.798-0.897], p < 0.0001). DWI hyperintensity in the putamen was a high-risk factor for deterioration during de-copper therapy (OR = 8.656, p < 0.05). The high-risk factors for readmission for intravenous de-copper therapies were DWI hyperintensity in the midbrain (OR = 3.818, p < 0.05) and the corpus callosum (OR = 2.654, p < 0.05). Both scoring systems had positive correlation with UWDRS scale (original semi-quantitative scoring system, r = 0.35, p < 0.001; consensus semi-quantitative scoring system, r = 0.351, p < 0.001.). Compared to the original scoring system, the consensus scoring system had higher correlations with the occurrence of deterioration (OR = 1.052, 95%CI [1.003, 1.0103], p < 0.05) and readmission for intravenous de-copper therapy (OR = 1.043, 95%CI [1.001, 1.086], p < 0.05). Conclusion: The predictive performance of the consensus semi-quantitative scoring system for cranial MRI was improved to guide medication, healthcare management, and prognosis prediction in patients with WD. For every point increase in the neuroimaging score, the risk of exacerbations during treatment increased by 5.2%, and the risk of readmission to the hospital within 6 months increased by 4.3%.

A Randomized Controlled Trial Evaluating the Effect of Tai Chi on the Drug Craving in Women.

This study was conducted in the purpose of investigating the effect of Tai Chi on drug craving for women with drug disorders. One hundred and twelve women were recruited from a drug rehabilitation center in China, and 47 and 48 were finally analyzed in the control group and exercise group, respectively. The exercise group underwent a 3-month Tai Chi training, whereas the control group experienced no exercise intervention during the same time period. The drug craving was measured by the visual analog scale. In data analysis, repeated-measures were utilized to test the differences between the control and exercise group over the course of the experiment time. The mean of the craving score significantly dropped from pre-test (control: mean = 5.38, SD = 3.04; exercise: mean = 4.68, SD = 2.93) to post-test (control: mean = 4.03, SD = 2.73; exercise: mean = 1.91, SD = 1.90) in both groups (control group: t = 3.84, df = 46, p < 0.001; exercise group: t = 5.941, df = 47, p < 0.001), with more decrease witnessed in the exercise group. Repeated-measures analysis with a Huynh-Feldt correction showed the significant effect of time (F = 27.383, p < 0.001) as well as the study group by time interaction (F = 3.52, p = 0.024). Tai Chi can ameliorate the drug craving in women and it could be a supportive treatment for drug addiction.

Effect of Eight-Month Exercise Intervention on Bone Outcomes of Young Opioid-Dependent Women.

OBJECTIVE: To evaluate the bone response to an 8 month aerobic gymnastics training program in young opioid-addicted women. DESIGN: Randomized controlled trial (parallel design). SETTING: Women's Specific Drug Rehabilitation Center in China. PATIENTS: One hundred and two young women with low bone quality and previous opioid addiction were divided into two groups: (a) the low bone quality intervention experimental group (n = 55; age: 30.3 ± 6.1) and (b) the low bone quality observed control group (observation group; n = 47; age: 29.0 ± 5.3). INTERVENTIONS: The intervention group took aerobic gymnastics regularly for 80 min/d and 5 d/wk for 8 months and completed follow-up testing. MAIN OUTCOME MEASURES: Substance use history and other life habits affecting bone quality were assessed by questionnaire-based interviews. Bone quality (stiffness-index, T-score, Z-score) was examined with quantitative ultrasound. Anthropometric characteristics (body weight, fat-free mass, fat mass) were obtained by bioelectrical impedance analysis. RESULTS: After the 8 month intervention, the stiffness index of bone quality increased significantly (before: 82 ± 6, after: 108 ± 14, p < 0.05) in the experimental group. However, the bone quality did not change significantly in the controls (before: 79 ± 10, after: 77 ± 13, p > 0.05). The bone change in the difference group was significant (experimental group: 31.7% vs observation group: -0.03%). Fat mass decreased in the experimental group (experimental group: before: 19.6 ± 3.7 kg, after: 18.8 ± 4.0 kg, p < 0.05). Meanwhile, the change in fat-free mass was the determination of the change in bone quality in the experimental group. CONCLUSIONS: Our results suggested that aerobic gymnastics intervention can be an effective strategy for the prevention and treatment of drug-induced osteoporosis in detoxification addicts.

Nutritional Status and Body Composition in Wilson Disease: A Cross-Sectional Study From China.

Background: Abnormal nutritional status is frequently seen in patients with chronic diseases. To date, no study has investigated the detailed characteristics of abnormal nutritional status among Wilson's disease (WD) patients in the Chinese cohort. This study aimed to describe the nutritional status of WD patients, with a particular focus on the differences between patients with different phenotypes. Methods: The study subjects comprised 119 healthy controls, 129 inpatients (hepatic subtype, n = 34; neurological subtype, n = 95) who were being treated at the affiliated hospital of the Institute of Neurology, Anhui University of Chinese Medicine. All of the subjects were assessed for body composition by using bioelectrical impedance analysis. All WD patients received anthropometry, nutritional risk screening 2002 (NRS2002), and laboratory test (hemocyte and serum biomarkers) additionally. Results: Compared with healthy controls, the fat mass and rate of total body and trunk were significantly higher in WD patients (P < 0.001), the muscle and skeletal muscle mass of total body and trunk were significantly lower in WD patients (P < 0.001). Compared with hepatic subtype patients, the fat mass and rate of total body, trunk, and limbs were significantly lower in neurological subtype patients (P<0.01); while there were no significant differences in muscle and skeletal muscle between these two subtypes. The overall prevalence of abnormal nutritional status in WD patients was 43.41% (56/129). The prevalence of high-nutritional risk and overweight in WD patients was 17.83% (23 of 129) and 25.58% (33 of 129), respectively. Compare with patients with high nutritional risk, macro platelet ratio, alkaline phosphatase, the basal metabolic rate (p < 0.05), creatinine, trunk fat rate (p < 0.01) and appendicular skeletal muscle mass (p < 0.001) were significantly higher in patients without nutritional risk (p < 0.001). Patients with a high nutritional risk tend to have a lower cholinesterase concentration (x 2 = 4.227, p < 0.05). Conclusion: Both patients with H-subtype and N-subtype are prone to have an abnormal nutritional status. Longitudinal studies are required to investigate if nutritional status and body composition could reflect prognosis in WD patients, and which of these body composition indexes contribute to malnutrition and worse prognosis.

Relative validity of an indirect calorimetry device for measuring resting energy expenditure and respiratory quotient.

BACKGROUND AND OBJECTIVES: Resting energy expenditure (REE) and respiratory quotient (RQ) are important for optimal nutritional care. The purpose of this study was to assess the accuracy and repeatability of an indirect calorimetry device (IIM-IC-100) in the measurement of REE and RQ in healthy Chinese adults. METHODS AND STUDY DESIGN: A total of 38 healthy adults (19 male and 19 female) aged 18-52 years (25±6 years) were enrolled in this study. REE and RQ were measured by IIM-IC-100 and by VO2000, alternately and in duplicate. RESULTS: There was a highly significant correlation between IIM-IC-100 REE and VO2000 REE (r=0.906, p<0.001), with mean IIM-IC-100 REE significantly higher than that of VO2000 (1475±269 vs 1394±313 kcal/d, p=0.002). Bland- Altman analysis revealed that the mean difference between IIM-IC-100 REE and VO2000 REE was 81.3 kcal/d, with limits of agreement of -185 to +347 kcal/d. There was no significant difference in RQ between the two devices. No significant differences were observed between the repeated measurements for both devices. Intrasubject coefficients of variation (CVs) of REE were smaller for IIM-IC-100 (5.8%) than for VO2000 (10.5%), while CVs of RQ were similar for IIM-IC-100 (7.2%) and VO2000 (6.9%). CONCLUSIONS: These preliminary data indicated that the IIM-IC-100 showed promise as an accurate and precise tool in the assessment of REE and RQ in healthy Chinese adults.

A comparison of bone quality and its determinants in young opioid-dependent women with healthy control group.

BACKGROUND: Little is known about bone quality and its determinants in patients with opioid addiction. The goal of this study was to compare bone quality and its determinants in young opioid addicted women with a local group of young healthy women. METHOD: Using cross-sectional design, 104 women (mean age 29.9 yrs, range: 20-40 yrs, SD=7.8) with previous opioid addiction and current methadone substitution (3-30mg, daily) for 1-16 weeks were compared to 117 healthy women (mean age 31.0 yrs, range: 20-40 yrs, SD=5.9). Bone quality was examined with quantitative ultrasound. Anthropometric characteristics (body weight, fat free mass (FFM), fat mass) were obtained by bioelectrical impedance analysis. Substance use and other risk factors for low bone quality were assessed by questionnaire-based interviews. RESULTS: More than one-quarter (34%) of patients had osteopenia (n=31) or osteoporosis (n=4), compared to 16% of the healthy control group having osteopenia (n=18). Bivariate correlation analysis demonstrated that age, body weight, and FFM correlated with bone quality (p<0.05) in healthy women, which were not found in patients. Multivariate analyses showed that in healthy controls, the determinants of bone quality were age, body height, physical activity, and BMI, but in patients, the determinant of bone quality was duration of drug intake. CONCLUSIONS: Long-term opioid dependence in young women may lead to low bone quality. Efforts to increase awareness of low bone quality in young opioid addicted women should be considered so that effective treatment may be employed to lower future fracture risk.

Correlation between ultrasonic power spectrum and bone density on the heel.

The purposes of this paper were to evaluate the correlation between ultrasonic power spectrum and bone density and to extract the effectiveness of parameters from power spectrum for evaluating bone density. A total of 50 persons 24-72years of age were recruited. All study participants underwent bone mineral density (BMD) measurements of the lumbar spine (vertebral levels L1-L4). The participants also underwent calcaneal measurements to determine ultrasonic power spectrum with central frequencies of 0.5MHz. Three parameters from normalized power spectrum, called principle frequency (PF), frequency band (FB), and amplitude for principle frequency (APF), were chosen and be evaluated the correlation with the lumbar spine BMD. The correlation coefficient of PF, FB and APF with BMD was r=-0.48 (p<0.001), r=0.48 (p<0.001), and r=-0.71 (p<0.001), respectively. The results showed that the correlation between APF and BMD was better than the correlation among PF, FB and BMD, and APF have a significant correlation with BMD. In conclusion, the correlations among the parameters of ultrasonic power spectrum and BMD are significant, and especially APF performs better than PF and FB in evaluating bone density of participants. These results suggest that ultrasonic power spectrum may contain substantial information not already contained in BUA and SOS. A multiple regression model including all three QUS variables was somewhat more predictive of BMD than a model including only BUA and SOS.

Quantitative ultrasound measurement of bone density based on dynamic time window: suitable for the measurement of speed of sound in radius.

PURPOSE: Our objective was to evaluate the dynamic time window method (DTWM) for the measurement of speed of sound (SOS) at the 1/3 distal radius in 10 human subjects in vivo. DTWM, which is based on our proprietary probe, is a new method for obtaining the accurate time of flight (TOF) of the first arriving signal (FAS). METHOD: This study compared DTWM with three conventional methods, i.e., threshold value method, first of maxima, and zero-crossing method, for the measurement of TOF of FAS. RESULTS: The highest correlation between SOS and bone mineral density (BMD) was found in the 10 human subjects using DTWM (R DTWM = 0.81, p < 0.005) rather than the three conventional methods (R = 0.37-0.41, p > 0.05). The individual short-term coefficient of variation (CV, from 0.07 to 0.27) and Root-Mean-Square (RMS) of the average CV (CVRMS, 0.18) were used to evaluate the reproducibility of our method (DTWM). CONCLUSION: Despite the small population, the results suggest that DTWM improves the accuracy of recognition of SOS in vivo, and might be applied to optimize the clinical assessment of long cortical bone osteoporosis and fracture risk profile.