Machine-learning models for activity class prediction: A comparative study of feature selection and classification algorithms

PurposeMachine-learning (ML) approaches have been repeatedly coupled with raw accelerometry to classify physical activity classes, but the features required to optimize their predictive performance are still unknown. Our aim was to identify appropriate combination of feature subsets and prediction algorithms for activity class prediction from hip-based raw acceleration data.MethodsThe hip-based raw acceleration data collected from 27 participants was split into training (70 %) and validation (30 %) subsets. A total of 206 time- (TD) and frequencydomain (FD) features were extracted from 6-second non-overlapping windows of the signal. Feature selection was done using seven filter-based, two wrapper-based, and one embedded algorithm, and classification was performed with artificial neural network (ANN), support vector machine (SVM), and random forest (RF). For every combination between the feature selection method and the classifiers, the most appropriate feature subsets were found and used for model training within the training set. These models were then validated with the left-out validation set.ResultsThe appropriate number of features for the ANN, SVM, and RF ranged from 20 to 45. Overall, the accuracy of all the three classifiers was higher when trained with feature subsets generated using filter-based methods compared with when they were trained with wrapper-based methods (range: 78.1 %–88 % vs. 66 %–83.5 %). TD features that reflect how signals vary around the mean, how they differ with one another, and how much and how often they change were more frequently selected via the feature selection methods.ConclusionsA subset of TD features from raw accelerometry could be sufficient for ML-based activity classification if properly selected from different axes.     

Ketorolac plus Lidocaine vs Lidocaine for pain relief following core needle soft tissue biopsy: A CONSORT-compliant double-blind randomized controlled study

Backgrounds:The main objective of this study was to compare the pain control efficacy of local administration of Lidocaine with or without the nonsteroidal anti-inflammatory drug, Ketorolac, and local conventional Lidocaine injection in core needle biopsy of the musculoskeletal tumor.Methods:The current study was a randomized, double-blind controlled clinical trial that included 128 patients with suspected musculoskeletal tumors. Patients were randomly assigned to either the Ketorolac plus Lidocaine (n = 64) or Lidocaine group (n = 64). The Ketorolac – Lidocaine combination syringe contained 30 mg Ketorolac and 2% Lidocaine – adrenaline dosage, and the Lidocaine syringe contained 2% Lidocaine – adrenaline dosage. The level of pain after core needle biopsy was evaluated for each patient at 1, 6, 12, 24, 48, and >48 hours by a Visual Analog Scale (VAS). The mean VAS changes over time were compared between the Ketorolac plus Lidocaine and Lidocaine groups using a linear mixed model.Results:baseline information including mean age of patients in Lidocaine group (51.5 ± 19.4 years) and in Lidocaine – Ketorolac combination group (50.1 ± 18 years), diagnosis (malignant, benign, metastatic, infection), tumor location (upper and lower extremities, back), VAS score 1-hour post-operation (mild and moderate pain) were noted. The VAS score ratings were significantly lower in Lidocaine – Ketorolac combination group when compared to the Lidocaine group during the 1 to 24 hours post-operation time period.Conclusion:Patients receiving Lidocaine – Ketorolac combination dosage had significantly lower VAS scores, and these results confirm that local injection of Lidocaine – Ketorolac combination had a superior pain-controlling effect during the first 24 hours after the biopsy procedure in comparison to Lidocaine injection alone, as measured by VAS score scale.     

Motor dysfunction in mild cognitive impairment as tested by kinematic analysis and transcranial magnetic stimulation

ObjectivePrevious studies have demonstrated voluntary movement alterations as well as motor cortex excitability and plasticity changes in patients with mild cognitive impairment (MCI). To investigate the pathophysiology of movement abnormalities in MCI, we tested possible relationships between movement abnormalities and primary motor cortex alterations in patients.MethodsFourteen amnestic MCI (aMCI) patients and 16 healthy controls were studied. Cognitive assessment was performed using clinical scales. Finger tapping was recorded by a motion analysis system. Transcranial magnetic stimulation was used to test the input/output curve of motor evoked potentials, intracortical inhibition, and short-latency afferent inhibition. Primary motor cortex plasticity was probed by theta burst stimulation. We investigated correlations between movement abnormalities, clinical scores, and cortical neurophysiological parameters.ResultsMCI patients showed less rhythmic movement but no other movement abnormalities. Cortical excitability measures were normal in patients, whereas plasticity was reduced. Movement rhythm abnormalities correlated with frontal dysfunction scores.ConclusionOur study in MCI patients demonstrated abnormal voluntary movement and plasticity changes, with no correlation between the two. Altered rhythm correlated with frontal dysfunction.SignificanceOur results contribute to the understanding of pathophysiological mechanisms of motor impairment in MCI.     

Development of an IgG-Fc fusion COVID-19 subunit vaccine,AKS-452

AKS-452 is a biologically-engineered vaccine comprising an Fc fusion protein of the SARS-CoV-2 viral spike protein receptor binding domain antigen (Ag) and human IgG1 Fc (SP/RBD-Fc) in clinical development for the induction and augmentation of neutralizing IgG titers against SARS-CoV-2 viral infection to address the COVID-19 pandemic. The Fc moiety is designed to enhance immunogenicity by increasing uptake via Fc-receptors (FcγR) on Ag-presenting cells (APCs) and prolonging exposure due to neonatal Fc receptor (FcRn) recycling. AKS-452 induced approximately 20-fold greater neutralizing IgG titers in mice relative to those induced by SP/RBD without the Fc moiety and induced comparable long-term neutralizing titers with a single dose vs. two doses. To further enhance immunogenicity, AKS-452 was evaluated in formulations containing a panel of adjuvants in which the water-in-oil adjuvant, Montanide™ ISA 720, enhanced neutralizing IgG titers by approximately 7-fold after one and two doses in mice, including the neutralization of live SARS-CoV-2 virus infection of VERO-E6 cells. Furthermore, ISA 720-adjuvanted AKS-452 was immunogenic in rabbits and non-human primates (NHPs) and protected from infection and clinical symptoms with live SARS-CoV-2 virus in NHPs (USA-WA1/2020 viral strain) and the K18 human ACE2-trangenic (K18-huACE2-Tg) mouse (South African B.1.351 viral variant). These preclinical studies support the initiation of Phase I clinical studies with adjuvanted AKS-452 with the expectation that this room-temperature stable, Fc-fusion subunit vaccine can be rapidly and inexpensively manufactured to provide billions of doses per year especially in regions where the cold-chain is difficult to maintain.     

高强度聚焦超声联合甲氨蝶呤在瘢痕子宫妊娠行宫腔镜微创清宫术中的应用效果分析

目的 探究采用高强度聚焦超声(HIFU)联合甲氨蝶呤(MTX)在瘢痕子宫妊娠行宫腔镜微创清宫术中的应用效果。方法 选择2018年5月—2019年5月在本院接受宫腔镜治疗的108例瘢痕子宫妊娠患者,依据随机数字表法分为研究组与对照组,每组各54例。对照组使用HIFU预处理,研究组则使用HIFU联合MTX的方式预处理,处理后两组均行宫腔镜辅助微创清宫术。通过随访分析比较两组的治疗效果及妊娠结局。结果 研究组术后7 d血清β-HCG水平为(28 598.38±2 176.52)U/L,低于对照组的(32 423.13±2 734.65)U/L,差异有统计学意义(t=8.042,P<0.001);研究组血清β-HCG转阴时间、月经恢复时间、住院时间分别为(21.33±3.78)d、(28.21±3.83)d、(6.53±2.12)d,分别短于对照组的(26.72±4.83)d、(33.82±4.91)d、(9.53±2.67)d,差异具有统计学意义(t=6.458、6.620、6.466,P均<0.001)。两组均随访≥24个月,宫内妊娠、未孕、流产、再次瘢痕子宫妊娠、异位妊娠比较,差异无统计学意义(χ2=1.895,P>0.05);宫内妊娠中足月分娩与早产比较,差异无统计学意义(χ2=0.131,P>0.05),研究组并发症发生率为22.58%,低于对照组的37.50%,差异有统计学意义(χ2=5.296,P=0.021)。结论 在宫腔镜辅助微创清宫术治疗瘢痕子宫妊娠前使用HIFU联合MTX预处理的效果优于单一使用HIFU预处理,可明显缩短患者血清β-HCG转阴时间、月经恢复时间与住院时间,但对女性的妊娠结局无影响,妇产科临床治疗前可结合患者需求择取最合适的预处理方案。     

On estimands arising from misspecified semiparametric rate-based analysis of recurrent episodic conditions

Marginal rate-based analyses are widely used for the analysis of recurrent events in clinical trials. In many areas of application, the events are not instantaneous but rather signal the onset of a symptomatic episode representing a recurrent infection, respiratory exacerbation, or bout of acute depression. In rate-based analyses, it is unclear how to best handle the time during which individuals are experiencing symptoms and hence are not at risk. We derive the limiting value of the Nelson-Aalen estimator and estimators of the regression coefficients under a semiparametric rate-based model in terms of an underlying two-state process. We investigate the impact of the distribution of the episode durations, heterogeneity, and dependence on the asymptotic and finite sample properties of standard estimators. We also consider the impact of these features on power in trials designed to test intervention effects on rate functions. An application to a trial of individuals with herpes simplex virus is given for illustration.     

Statin use and medically attended acute respiratory illness among influenza vaccine recipients

BackgroundPrevious studies have suggested that statins decrease influenza vaccine effectiveness and increase risk of medically attended acute respiratory illness (MAARI).ObjectivesTo examine the association of incident statin use and MAARI in a cohort of influenza vaccine recipients.MethodsThis retrospective cohort study evaluated influenza vaccine recipients within the Tricare population. The primary outcome compared MAARI incidence during the follow-up period in a propensity score-matched cohort of incident statin users and statin non-users. Secondary analysis included propensity score-adjusted comparisons between incident statin users and statin non-users in the entire cohort and prespecified sub-cohorts with and without comorbidities. The propensity score was derived from 72 variables encompassing demographics, medical history, comorbidities, medication use, and healthcare utilization.ResultsMAARI incidence in statin users was similar to non-users in the propensity score-matched cohort (odds ratio [OR] 0.92; 95% confidence interval [CI] 0.84–1.01). In contrast, statin users with lower comorbidity had lower OR for MAARI compared to non-users (Charlson Score zero cohort: 0.85 [CI 0.74–0.98]; No Diabetes cohort: 0.88 [CI 0.80–0.96]).ConclusionIncident statin use was not associated with increased MAARI incidence and may be associated with lower incidence of MAARI in those with less comorbidity. This study thus offers reassurance regarding the effectiveness of the influenza vaccine in statin users.