A Multi-Label Based Physical Activity Recognition via Cascade Classifier.

Physical activity recognition is a field that infers human activities used in machine learning techniques through wearable devices and embedded inertial sensors of smartphones. It has gained much research significance and promising prospects in the fields of medical rehabilitation and fitness management. Generally, datasets with different wearable sensors and activity labels are used to train machine learning models, and most research has achieved satisfactory performance for these datasets. However, most of the methods are incapable of recognizing the complex physical activity of free living. To address the issue, we propose a cascade classifier structure for sensor-based physical activity recognition from a multi-dimensional perspective, with two types of labels that work together to represent an exact type of activity. This approach employed the cascade classifier structure based on a multi-label system (Cascade Classifier on Multi-label, CCM). The labels reflecting the activity intensity would be classified first. Then, the data flow is divided into the corresponding activity type classifier according to the output of the pre-layer prediction. The dataset of 110 participants has been collected for the experiment on PA recognition. Compared with the typical machine learning algorithms of Random Forest (RF), Sequential Minimal Optimization (SMO) and K Nearest Neighbors (KNN), the proposed method greatly improves the overall recognition accuracy of ten physical activities. The results show that the RF-CCM classifier has achieved 93.94% higher accuracy than the 87.93% obtained from the non-CCM system, which could obtain better generalization performance. The comparison results reveal that the novel CCM system proposed is more effective and stable in physical activity recognition than the conventional classification methods.

Tofacitinib for new-onset adult patients with anti-melanoma differentiation-associated 5 gene antibody positive dermatomyositis.

OBJECTIVE: We aimed to investigate the efficacy and safety of tofacitinib in adult anti-melanoma differentiation-associated 5 gene (Anti-MDA5) antibody-positive dermatomyositis (DM) patients and evaluate the effects of tofacitinib on peripheral lymphocyte subsets. METHODS: An open-label study was conducted of 15 new-onset, untreated adult patients with anti-MDA5-positive DM for tofacitinib with a dose of 5mg twice per day. The primary outcome was defined by the total improvement score after treatment for 6 months, classified according to the 2016 American College of Rheumatology/ European League Against Rheumatism (ACR/EULAR) response criteria for adult DM and polymyositis. Secondary outcomes after 6 months treatment comprised the change in predicted forced vital capacity, the percentage of predicted carbon monoxide diffusion capacity, ferritin level and peripheral blood lymphocyte subsets measured by flow cytometry. RESULTS: Disease responses occurred in 10 patients (71.4%) after 6 months. The median total improvement score was 43.75 (41.875-59.375). Two patients achieved major improvement, seven achieved moderate and one minimal. The serum ferritin level (p = 0.008), DLCO% (p = 0.009) was improved and a marked increase in total lymphocyte cells (p = 0.045) and CD8+ T cells (p = 0.006) was measured after 6 months treatment compared to baseline. CONCLUSION: Tofacitinib demonstrates efficacy for new-onset, untreated adult patients with anti-MDA5-positive DM and stimulates proliferation of peripheral lymphocyte subsets (especially total lymphocyte cells and CD8+ T cells) after 6 months treatment. Further studies are warranted to validate the current findings. Key Points • Treatment of anti-melanoma differentiation-associated 5 gene antibody positive dermatomyositis is always challenging. • This prospective, open-label clinical trial demonstrates tofacitinib is an effective and safe agent for new-onset adult patients with anti-MDA5-positive DM. • Tofacitinib treatment results in an increase in peripheral lymphocyte numbers, especially CD8+ T cells at 6 months compared with pre-treatment levels.

The rapid inhibition of B-cell activation markers by belimumab was associated with disease control in systemic lupus erythematosus patients.

Objective: To examine the kinetics of B cell subsets and activation markers in the early stage of belimumab treatment and their correction with treatment response. Methods: We enrolled 27 systemic lupus erythematosus (SLE) patients receiving 6 months belimumab treatment. Flow cytometry was used to test their B cell subsets and activation markers (including CD40, CD80, CD95, CD21low, CD22, p-SYK and p-AKT). Results: During belimumab treatment, SLEDAI-2K declined, the proportions of CD19+ B cells and naïve B cells decreased, whereas the switched memory B cells and non-switched B cells increased. The larger variations of the B cell subsets and the activation markers were in the first 1 month than the other later time frames. The ratio of p-SYK/p-AKT on non-switched B cell at 1 month was associated with the SLEDAI-2K decline rate in the 6 months of belimumab treatment. Conclusion: B cell hyperactivity was rapidly inhibited in the early stage of belimumab treatment, and the ratio of p-SYK/p-AKT may predict SLEDAI-2K decline. Clinical Trial Registration: https://www.clinicaltrials.gov/ct2/show/NCT04893161?term=NCT04893161&draw=2&rank=1; identifier: NCT04893161.

EvtSNN: Event-driven SNN simulator optimized by population and pre-filtering.

Recently, spiking neural networks (SNNs) have been widely studied by researchers due to their biological interpretability and potential application of low power consumption. However, the traditional clock-driven simulators have the problem that the accuracy is limited by the time-step and the lateral inhibition failure. To address this issue, we introduce EvtSNN (Event SNN), a faster SNN event-driven simulator inspired by EDHA (Event-Driven High Accuracy). Two innovations are proposed to accelerate the calculation of event-driven neurons. Firstly, the intermediate results can be reused in population computing without repeated calculations. Secondly, unnecessary peak calculations will be skipped according to a condition. In the MNIST classification task, EvtSNN took 56 s to complete one epoch of unsupervised training and achieved 89.56% accuracy, while EDHA takes 642 s. In the benchmark experiments, the simulation speed of EvtSNN is 2.9-14.0 times that of EDHA under different network scales.

Low positivity rates for HBeAg and HBV DNA in rheumatoid arthritis patients: a case-control study.

BACKGROUND: The rates of hepatitis B virus (HBV) infection in rheumatoid arthritis (RA) patients are controversial when considering the reported outcomes. It was speculated that HBV infection status was altered after RA, and variations inn HBV infection rates became apparent. METHODS: To compare the positive proportions of hepatitis B e antigen (HBeAg) and HBV DNA, a retrospective case-control study was performed between 27 chronic hepatitis B (CHB) patients with RA and 108 age- and gender-matched CHB patients. In addition, the positivity rates of hepatitis B surface antigen (HBsAg) and hepatitis B core antibody (anti-HBc) were surveyed among the 892 RA patients. RESULTS: Compared to CHB patients, CHB patients with RA exhibited lower rates of HBeAg positivity (11.1% vs. 35.2%, P = 0.003), HBV DNA positivity (37.0% vs. 63.9%, P = 0.007) and ALT elevation (11.1% vs. 35.2%, P = 0.024). In the 892 RA patients, the prevalence of HBsAg (3.0%) was lower than that reported in the Chinese national data (7.2%), whereas the anti-HBc positivity rate of 44.6% was higher than that of 34.1%. CONCLUSION: HBV infection status was altered after suffering from RA. Compared to the matched CHB patients, low positive proportions of HBeAg and HBV DNA were observed for CHB patients with RA.

ALSA: Associative Learning Based Supervised Learning Algorithm for SNN.

Spiking neural network (SNN) is considered to be the brain-like model that best conforms to the biological mechanism of the brain. Due to the non-differentiability of the spike, the training method of SNNs is still incomplete. This paper proposes a supervised learning method for SNNs based on associative learning: ALSA. The method is based on the associative learning mechanism, and its realization is similar to the animal conditioned reflex process, with strong physiological plausibility and rationality. This method uses improved spike-timing-dependent plasticity (STDP) rules, combined with a teacher layer to induct spikes of neurons, to strengthen synaptic connections between input spike patterns and specified output neurons, and weaken synaptic connections between unrelated patterns and unrelated output neurons. Based on ALSA, this paper also completed the supervised learning classification tasks of the IRIS dataset and the MNIST dataset, and achieved 95.7 and 91.58% recognition accuracy, respectively, which fully proves that ALSA is a feasible SNNs supervised learning method. The innovation of this paper is to establish a biological plausible supervised learning method for SNNs, which is based on the STDP learning rules and the associative learning mechanism that exists widely in animal training.

Pruning Growing Self-Organizing Map Network for Human Physical Activity Identification.

Human physical activity identification based on wearable sensors is of great significance to human health analysis. A large number of machine learning models have been applied to human physical activity identification and achieved remarkable results. However, most human physical activity identification models can only be trained based on labeled data, and it is difficult to obtain enough labeled data, which leads to weak generalization ability of the model. A Pruning Growing SOM model is proposed in this paper to address the limitations of small-scale labeled dataset, which is unsupervised in the training stage, and then only a small amount of labeled data is used for labeling neurons to reduce dependency on labeled data. In training stage, the inactive neurons in network can be deleted by pruning mechanism, which makes the model more consistent with the data distribution and improves the identification accuracy even on unbalanced dataset, especially for the action categories with poor identification effect. In addition, the pruning mechanism can also speed up the inference of the model by controlling its scale.

QTc interval prolongation in the patients with primary biliary cholangitis.

BACKGROUND: The QT interval prolongation was associated with fatal arrhythmias and cardiac death. However, there were not adequate data to clarify the situation of QT interval prolongation in primary biliary cholangitis (PBC) patients. The aim of this study was to clarify the rate and the associated risk factors of corrected QT (QTc) interval prolongation in PBC patients. METHODS: From January 2016 to December 2020, PBC patients were retrospectively enrolled. The rate of QTc interval prolongation was surveyed and the associated risk factors were clarified by univariate and multivariate analyses. RESULTS: Among the 189 PBC patients, 24.3% (46/189) had the QTc interval prolongation. The univariate analysis showed that age, Child-Pugh classification, creatinine, international normalized ratio (INR), and platelet (PLT) were associated with QTc interval prolongation in the PBC patients. The multivariate analysis further showed only age (p = .028) and Child-Pugh classification (p = .035) were the associated risk factors. It had the highest risk of QTc interval prolongation (as high as 64.3%) in the patients who were more than 62.5 years old and with Child-Pugh C. CONCLUSION: The QTc interval prolongation was frequent in PBC patients, especially in the patients with decompensated cirrhosis. The rate of QTc interval prolongation was as high as 64.3% in the PBC patients who were more than 62.5 years old and classified as Child-Pugh C.

Elevated CD19+Siglec-10+ B cell levels are correlated with systemic lupus erythematosus disease activity.

BACKGROUND: Systemic lupus erythematosus (SLE) is a chronic autoimmune disease characterized by B cell dysregulation and the breakdown of self-tolerance, leading to pathogenic autoantibody production. Human Siglec-10 is a member of the sialic acid-binding immunoglobulin-type lectin (Siglec) family and a B cell surface coreceptor that inhibits B cell receptor-induced signalling. However, to date, no report has investigated CD19+Siglec-10+ B cells in SLE patients. Thus, this study aimed to measure the population of CD19+Siglec-10+ B cells in patients with SLE and its correlation with disease activity. METHODS: Flow cytometry was employed to measure the population of CD19+Siglec-10+ B cells in peripheral blood mononuclear cells (PBMCs) of both SLE patients and healthy controls (HCs). The correlation of the proportion of CD19+Siglec-10+ B cells with the values of SLE disease activity was analysed. PBMCs from HCs were challenged with serum from active SLE, inactive SLE, or HCs, and the proportion of CD19+Siglec-10+ B cells was then assessed. The effect of dexamethasone (DEX) or hydroxychloroquine (HCQ) treatment on the proportion of CD19+Siglec-10+ B cells in PBMCs from SLE patients was also determined. RESULTS: The proportion of CD19+Siglec-10+ B cells in SLE patients was significantly elevated (P < 0.05), correlated positively with the SLEDAI score (r = 0.304; P = 0.018) and negatively with complement component 3 (C3) (r = -0.283; P = 0.04). In vitro assays indicated that sera from active SLE patients could significantly enhance the proportion of CD19+Siglec-10+ B cells (P < 0.05), while HCQ treatment significantly attenuated their proportions (P < 0.01). CONCLUSIONS: The elevation of CD19+Siglec-10+ B cells and their correlation with disease activity may suggest a role for Siglec-10 in the pathogenesis and progression of SLE and provide a serum biomarker for SLE activity.

Assessment of Low Bone Mineral Density in Untreated Patients with Takayasu's Arteritis.

Chronic inflammation affects bone metabolism and accelerates bone loss. This study is aimed at analyzing the prevalence of low bone mineral density (LBMD) in patients with untreated Takayasu's arteritis (TA) and risk factors. Forty untreated TA patients were enrolled, including 38 premenopausal women and 2 men before 50 years old. The control group included 60 age- and gender-matched healthy persons. Bone mineral density (BMD) of lumbar vertebrae and hip in patients with TA and the control group was measured by the dual-energy X-ray method. Serum 25OHD and ß-CTX were also measured. The lumbar BMD of TA patients (0.89 ± 0.11 g/cm2) was significantly lower than that of the healthy control (0.97 ± 0.11 g/cm2). The prevalence of LBMD at the lumbar spine (17.50%) was significantly higher than that of the control group (3.33%). However, there was no significant difference at the hip. The 25OHD of TA patients was lower than that of healthy controls, while the level of ß-CTX was higher. The levels of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) in patients with LBMD were higher than those in patients with normal BMD. According to univariate correlation analysis, there was a significant negative correlation between LDL-C and lumbar BMD. Binary logistic regression analysis showed that LDL-C was an important factor affecting the occurrence of LBMD in patients with TA (OR = 25.269, P = 0.02). Our result reveals bone loss in TA patients, which hints the relationship among inflammation, lipid metabolism, and bone metabolism.

Loss of microRNA-147 function alleviates synovial inflammation through ZNF148 in rheumatoid and experimental arthritis.

MicroRNA-147 (miR-147) had been previously found induced in synoviocytes by inflammatory stimuli derived from T cells in experimental arthritis. This study was designed to verify whether loss of its function might alleviate inflammatory events in joints of experimental and rheumatoid arthritis (RA). Dark Agouti (DA) rats were injected intradermally with pristane to induce arthritis, and rno-miR-147 antagomir was locally administrated into individual ankle compared with negative control or rno-miR-155-5p antagomir (potential positive control). Arthritis onset, macroscopic severity, and pathological changes were monitored. While in vitro, gain or loss function of hsa-miR-147b-3p/hsa-miR-155-5p and ZNF148 was achieved in human synovial fibroblast cell line SW982 and RA synovial fibroblasts (RASF). The expression of miRNAs and mRNAs was detected by using RT-quantitative PCR, and protein expression was detected by using Western blotting. Anti-miR-147 therapy could alleviate the severity, especially for the synovitis and joint destruction in experimental arthritis. Gain of hsa-miR-147b-3p/hsa-miR-155-5p function in TNF-α stimulated SW982 and RASF cells could upregulate, in contrast, loss of hsa-miR-147b-3p/hsa-miR-155-5p function could downregulate the gene expression of TNF-α, IL-6, MMP3, and MMP13. Hence, such alteration could participate in synovial inflammation and joint destruction. RNAi of ZNF148, a miR-147's target, increased gene expression of TNF-α, IL-6, MMP3, and MMP13 in SW982 and RASF cells. Also, mRNA sequencing data showed that hsa-miR-147b-3p mimic and ZNF148 siRNA commonly regulated the gene expression of CCL3 and DEPTOR as well as some arthritis and inflammation-related pathways. Taken together, miR-147b-3p contributes to synovial inflammation through repressing ZNF148 in RA and experimental arthritis.

Clustering Ensemble Model Based on Self-Organizing Map Network.

This paper proposes a clustering ensemble method that introduces cascade structure into the self-organizing map (SOM) to solve the problem of the poor performance of a single clusterer. Cascaded SOM is an extension of classical SOM combined with the cascaded structure. The method combines the outputs of multiple SOM networks in a cascaded manner using them as an input to another SOM network. It also utilizes the characteristic of high-dimensional data insensitivity to changes in the values of a small number of dimensions to achieve the effect of ignoring part of the SOM network error output. Since the initial parameters of the SOM network and the sample training order are randomly generated, the model does not need to provide different training samples for each SOM network to generate a differentiated SOM clusterer. After testing on several classical datasets, the experimental results show that the model can effectively improve the accuracy of pattern recognition by 4%∼10%.

Effect of 1,25-(OH)2D3 on Proliferation of Fibroblast-Like Synoviocytes and Expressions of Pro-Inflammatory Cytokines through Regulating MicroRNA-22 in a Rat Model of Rheumatoid Arthritis.

OBJECTIVE: This study aims to investigate the regulatory mechanism of 1,25-(OH)2D3 on the proliferation of fibroblast-like synoviocytes (FLS) and expressions of pro-inflammatory cytokines in rheumatoid arthritis (RA) rats via microRNA-22 (miR-22). METHODS: A rat model of RA was established with a subcutaneous injection of type II collagen. After treated with different concentrations of 1,25-(OH)2D3 the proliferation of FLS was estimated by the MTT method, and the optimal concentration of 1,25-(OH)2D3 was selected for further experiments. Cell proliferation was detected by MTT. Cell cycle and apoptosis were analyzed by FCM. The IL-1ß, IL-6, IL-8, and PGE2 protein expressions were determined by ELISA, and MMP-3, INOS, and Cox-2 mRNA expressions were measured by qRT-PCR. RESULTS: The rat model of RA was successfully established. Compared with the blank group, the 1,25-(OH)2D3 and miR-22 inhibitors groups exhibited higher proliferation inhibition and apoptosis rates, lower levels of pro-inflammatory cytokines (IL-1ß, IL-6, IL-8, and PGE2), and decreased mRNA expressions of MMP-3, INOS, and Cox-2. The miR-22 mimics group had lower proliferation inhibition and apoptosis rates, elevated expressions of pro-inflammatory cytokines and MMP-3, INOS, and Cox-2 than the blank group. In contrast to the 1,25-(OH)2D3 group, the proliferation inhibition and apoptosis rates were down-regulated, and the expressions of pro-inflammatory cytokines and MMP-3, INOS, and Cox-2 were up-regulated in the 1,25-(OH)2D3 + miR-22 mimics group. CONCLUSION: Our study demonstrated that 1,25-(OH)2D3 inhibits the proliferation of FLS and alleviates inflammatory response in RA rats by down-regulating miR-22.

Association of Fcgamma receptor type 2A and 3A genotypes with rheumatoid arthritis in Chinese population.

AIM: SNPs of FcγRs were implicated in pathogenesis of rheumatoid arthritis (RA) and treatment efficacy of TNF inhibitors (TNFi). This study aims to investigate the associations of FcγRIIa and FcγRIIIa genotypes with autoantibody production and treatment response to TNFi in Chinese patients with RA. PATIENTS & METHODS: FcγRIIa and FcγRIIIa polymorphisms were genotyped in 158 RA patients. Response to TNFi was evaluated in 18 patients at 3 and 6 months after treatment. RESULTS: FcγRIIa-131H allele was significantly increased in autoantibody-negative RA patients. FcγRIIa-131H/H+H/R was closely associated with differences in 28-joint disease activity score in patients at months 3 and 6 of TNFi treatment. CONCLUSION: FcγRIIa-131H allele may have a protective role in autoantibody production and might be a biomarker for predicting good response to TNFi in Chinese RA patients.

A Random Forest-based ensemble method for activity recognition.

This paper presents a multi-sensor ensemble approach to human physical activity (PA) recognition, using random forest. We designed an ensemble learning algorithm, which integrates several independent Random Forest classifiers based on different sensor feature sets to build a more stable, more accurate and faster classifier for human activity recognition. To evaluate the algorithm, PA data collected from the PAMAP (Physical Activity Monitoring for Aging People), which is a standard, publicly available database, was utilized to train and test. The experimental results show that the algorithm is able to correctly recognize 19 PA types with an accuracy of 93.44%, while the training is faster than others. The ensemble classifier system based on the RF (Random Forest) algorithm can achieve high recognition accuracy and fast calculation.

Preliminary exploration of the measurement of walking speed for the apoplectic people based on UHF RFID.

The number of the apoplectic people is increasing while population aging is quickening its own pace. The precise measurement of walking speed is very important to the rehabilitation guidance of the apoplectic people. The precision of traditional measuring methods on speed such as stopwatch is relatively low, and high precision measurement instruments because of the high cost cannot be used widely. What's more, these methods have difficulty in measuring the walking speed of the apoplectic people accurately. UHF RFID tag has the advantages of small volume, low price, long reading distance etc, and as a wearable sensor, it is suitable to measure walking speed accurately for the apoplectic people. In order to measure the human walking speed, this paper uses four reader antennas with a certain distance to reads the signal strength of RFID tag. Because RFID tag has different RSSI (Received Signal Strength Indicator) in different distances away from the reader, researches on the changes of RSSI with time have been done by this paper to calculate walking speed. The verification results show that the precise measurement of walking speed can be realized by signal processing method with Gaussian Fitting-Kalman Filter. Depending on the variance of walking speed, doctors can predict the rehabilitation training result of the apoplectic people and give the appropriate rehabilitation guidance.

ANCA associated glomerulonephritis in a patient with mixed connective tissue disease.

OBJECTIVE: To investigate the diagnosis and treatment of mixed connective tissue disease (MCTD) and myeloperoxidase-antineutrophil cytoplasmic antibody (MPO-ANCA) associated vasculitis, which is a rare clinical entity in medical practice. METHODS: A 35-year-old female of Asian origin was admitted to our hospital due to complaints of Raynaud's phenomenon, myalgia, arthralgia and fatigue. The patient was diagnosed as MCTD in the out-patient department 8 months prior to admission based on Alarcon-Segovia classification criteria of Raynaud's phenomenon, myalgia, arthralgia and a high anti-U1 ribonucleoprotein antibody level. Interstitial lung disease was determined by chest computed tomography. Renal biopsy was performed because of marked proteinuria on 24 h urine collection. Histopathological examination revealed glomerulonephritis with fibrocellular/cellular crescents, in which moderate staining of IgM was shown by direct immunofluorescence. She was tested positive for myeloperoxidase antineutrophil cytoplasmic antibody. RESULTS: High dose of methylprednisolone (500 mg/d for 3 days) was started intravenously when the results of renal biopsy were obtained. Oral prednisone and intravenous cyclophosphamide therapy (0.8 g/month) were continued for 12 months. Daily urinary protein loss decreased dramatically and serum creatinine was maintained at a normal level. CONCLUSION: Corticosteroids and cyclophosphamide are effective in the treatment of MPO-ANCA associated glomerulonephritis in MCTD.

Wireless design of a multisensor system for physical activity monitoring.

Real-time monitoring of human physical activity (PA) is important for assessing the intensity of activity and exposure to environmental pollutions. A wireless wearable multisenor integrated measurement system (WIMS) has been designed for real-time measurement of the energy expenditure and breathing volume of human subjects under free-living conditions. To address challenges posted by the limited battery life and data synchronization requirement among multiple sensors in the system, the ZigBee communication platform has been explored for energy-efficient design. Two algorithms have been developed (multiData packaging and slot-data-synchronization) and coded into a microcontroller (MCU)-based sensor circuitry for real-time control of wireless data communication. Experiments have shown that the design enables continued operation of the wearable system for up to 68 h, with the maximum error for data synchronization among the various sensor nodes (SNs) being less than 24 ms. Experiment under free-living conditions have shown that the WIMS is able to correctly recognize the activity intensity level 86% of the time. The results demonstrate the effectiveness of the energy-efficient wireless design for human PA monitoring.

ZigBee-based wireless multi-sensor system for physical activity assessment.

Physical activity (PA) is important for assessing human exposure to the environment. This paper presents a ZigBee-based Wireless wearable multi-sensor Integrated Measurement System (WIMS) for in-situ PA measurement. Two accelerometers, a piezoelectric displacement sensor, and an ultraviolet (UV) sensor have been used for the physical activity assessment. Detailed analysis was performed for the hardware design and embedded program control, enabling efficient data sampling and transmission, compact design, and extended battery life to meet requirements for PA assessment under free-living conditions. Preliminary testing of the WIMS has demonstrated the functionality of the design, while performance comparison of the WIMS with a wired version on an electromagnetic shaker has demonstrated the signal validity.