Residual-based tree-structured survival analysis.

Extensions of various non-parametric regression techniques (for example, additive models, trees, MARS) have been devised for right censored survival data. These approaches directly handle the difficulties posed by censoring. However, it is possible to bypass these difficulties by utilizing standard non-parametric regression procedures applied with (say) martingale residuals as outcome. Analytic correspondences between the direct and residual-based approaches have been established for additive models while more qualitative comparisons have been provided for MARS. Here we develop such correspondences for tree-structured regression. In particular, we provide an analytic relationship between logrank and martingale residual sum-of-squares split functions that explains the widely observed similarity of the resultant trees. Further investigation is provided by simulation and an illustrative example using time to AIDS with data deriving from a Western Australian HIV cohort study.     

A comparison of regression trees, logistic regression, generalized additive models, and multivariate adaptive regression splines for predicting AMI mortality

Clinicians and health service researchers are frequently interested in predicting patient-specific probabilities of adverse events (e.g. death, disease recurrence, post-operative complications, hospital readmission). There is an increasing interest in the use of classification and regression trees (CART) for predicting outcomes in clinical studies. We compared the predictive accuracy of logistic regression with that of regression trees for predicting mortality after hospitalization with an acute myocardial infarction (AMI). We also examined the predictive ability of two other types of data-driven models: generalized additive models (GAMs) and multivariate adaptive regression splines (MARS). We used data on 9484 patients admitted to hospital with an AMI in Ontario. We used repeated split-sample validation: the data were randomly divided into derivation and validation samples. Predictive models were estimated using the derivation sample and the predictive accuracy of the resultant model was assessed using the area under the receiver operating characteristic (ROC) curve in the validation sample. This process was repeated 1000 times-the initial data set was randomly divided into derivation and validation samples 1000 times, and the predictive accuracy of each method was assessed each time. The mean ROC curve area for the regression tree models in the 1000 derivation samples was 0.762, while the mean ROC curve area of a simple logistic regression model was 0.845. The mean ROC curve areas for the other methods ranged from a low of 0.831 to a high of 0.851. Our study shows that regression trees do not perform as well as logistic regression for predicting mortality following AMI. However, the logistic regression model had performance comparable to that of more flexible, data-driven models such as GAMs and MARS.     

Modeling of air pollutants using least square support vector regression,multivariate adaptive regression spline,and M5 model tree models

This study investigates the applicability of three different soft computing methods, least square support vector regression (LSSVR), multivariate adaptive regression splines (MARS), and M5 Model Tree (M5-Tree), in forecasting SO₂ concentration. These models were applied to monthly data obtained from Janakpuri, Nizamuddin, and Shahzadabad, located in Delhi, India. The models were compared with each other using the cross validation method with respect to root mean square error, mean absolute error, and correlation coefficient. According to the comparison, LSSVR provided better accuracy than the other models, while the MARS model was found to be the second best model in forecasting monthly SO₂ concentration. Results indicated that the applied models gave better forecasting accuracy in Janakpuri station than the other stations. The results were also compared with previous studies and satisfactory results were obtained from three methods in modeling SO₂ concentrations.     

Multivariate adaptive regression splines: a powerful method for detecting disease-risk relationship differences among subgroups

In a wide variety of medical research scenarios one is interested in the question whether regression curves differ for subgroups in the sample. Examples are gender differences in the effect of drug treatment or the study of genotype-environment interactions. To address this question exploratory techniques are often required because detailed knowledge concerning the shape of the regression curves and how that shape differs across subgroups is lacking. In this article we explored the power of two such exploratory techniques: multivariate adaptive regression splines (MARS) and least squares curve fitting using polynomials. For this purpose simulations were performed using linear, logistic, and complex non-linear curves. The power obtained from MARS was on average 1.4 times higher than with polynomials. It was shown that power was higher even if the regression curve was linear, that gains increased with the complexity of the curve, and that for highly non-linear curves model-free methods such as MARS might be the only alternative.     

Performance of variable and function selection methods for estimating the nonlinear health effects of correlated chemical mixtures: A simulation study

Statistical methods for identifying harmful chemicals in a correlated mixture often assume linearity in exposure-response relationships. Nonmonotonic relationships are increasingly recognized (eg, for endocrine-disrupting chemicals); however, the impact of nonmonotonicity on exposure selection has not been evaluated. In a simulation study, we assessed the performance of Bayesian kernel machine regression (BKMR), Bayesian additive regression trees (BART), Bayesian structured additive regression with spike-slab priors (BSTARSS), generalized additive models with double penalty (GAMDP) and thin plate shrinkage smoothers (GAMTS), multivariate adaptive regression splines (MARS), and lasso penalized regression. We simulated realistic exposure data based on pregnancy exposure to 17 phthalates and phenols in the US National Health and Nutrition Examination Survey using a multivariate copula. We simulated data sets of size N = 250 and compared methods across 32 scenarios, varying by model size and sparsity, signal-to-noise ratio, correlation structure, and exposure-response relationship shapes. We compared methods in terms of their sensitivity, specificity, and estimation accuracy. In most scenarios, BKMR, BSTARSS, GAMDP, and GAMTS achieved moderate to high sensitivity (0.52-0.98) and specificity (0.21-0.99). BART and MARS achieved high specificity (≥0.90), but low sensitivity in low signal-to-noise ratio scenarios (0.20-0.51). Lasso was highly sensitive (0.71-0.99), except for quadratic relationships (≤0.27). Penalized regression methods that assume linearity, such as lasso, may not be suitable for studies of environmental chemicals hypothesized to have nonmonotonic relationships with outcomes. Instead, BKMR, BSTARSS, GAMDP, and GAMTS are attractive methods for flexibly estimating the shapes of exposure-response relationships and selecting among correlated exposures.     

Tree and spline based association analysis of gene-gene interaction models for ischemic stroke

In the biology of complex disorders, such as atherothrombosis, interactions among genetic factors may play an important role, and theoretical considerations suggest that gene-gene interactions are quite common in such diseases. We used a nested case-control sample from the Physicians' Health Study, a randomized trial assessing the effects of aspirin and beta-carotene on cardiovascular disease and cancer among 22071 US male physicians, to examine these relationships for ischemic stroke. Data were available on 92 polymorphisms from 56 candidate genes related to inflammation, thrombosis and lipid metabolism, assessed in 319 incident cases of ischemic stroke and 2090 disease-free controls. We used classification and regression trees (CART) and multivariate adaptive regression spline (MARS) models to explore the presence of genetic interactions in these data. These models offer advantages over typical logistic regression methods in that they may uncover interactions among genes that do not exhibit strong marginal effects. Final models were selected using either the Bayes Information Criterion or cross-validation. Model fit was assessed using 10-fold cross-validation of the entire selection process. Both the CART and two-way MARS-logit models identified an interaction between two polymorphisms linked to inflammation, the P-selectin (val640leu) and interleukin-4 (C(582) T) genes. Internal validation of these models, however, suggested that effects of these polymorphisms are additive. Although further external validation of these models is necessary, these methods may be valuable in exploring and identifying potential gene-gene as well as gene-environment interactions in association studies.     

Common Disease Analysis Using Multivariate Adaptive Regression Splines (MARS): Genetic Analysis Workshop 12 Simulated Sequence Data

A newly developed modern analytic approach, Multivariate Adaptive Regression Splines (MARS), was used to identify both genetic and non‐genetic factors involved in the etiology of a common disease. We tested this method on the simulated data provided by the Genetic Analysis Workshop (GAW) 12 in problem 2 for the isolated population. MARS simultaneously analyzes all inputs, in this case DNA sequence variants and non‐genetic data, and selectively prunes away variables contributing insignificantly to fit by internal cross‐validation to arrive at a generalizable predictive model of the response. The relevant factors identified, by means of an importance value computed by MARS, were assumed to be associated with risk to the disease. The application of a series of subsequent models identified the quantitative traits and a single major gene contributing directly to risk liability using five sets of 7,000 individuals. © 2001 Wiley‐Liss, Inc.     

Management of prenatal care information: integration of the problem list and clinical comments

To improve the functionality of our MAternal Record System (MARS), we reviewed the clinical problems entered into the MARS database from approximately 2800 patient records and interviewed representative clinicians about their use of the Clinical Comments and Problem List sections of MARS. Problems were assigned a unique term that corrected for spelling, spacing, synonyms, and abbreviation variations. Analysis of these terms suggested design changes that would increase the number of unique entries into the Problem List, including (1) modifications in the automated problem entry functions within MARS and (2) integration of the Problem List and Clinical Comments sections.     

不明位置妊娠Logistic回归超声诊断模型的初步研究

目的:通过经阴道超声(TVS)检测不明位置妊娠(PUL)患者的子宫内膜,建立鉴别诊断的多变量回归模型,并评价该模型的临床应用价值。方法:分析早期PUL患者的子宫内膜二维声像图特征,并对年龄、β-HCG、停经时间等进行分析,采用多变量Logistic回归法及ROC曲线筛选出有意义的诊断指标,建立PUL患者的多变量回归模型。结果:建立了三个预测PUL妊娠结局的Logistic回归模型,分别对宫内早孕(MIUP)、宫内早孕流产(MIUPM)和早期异位妊娠(MEP)进行评估,ROC曲线下面积(AUC)分别为0.941、0.886、0.891,预测模型循证试验的诊断指标均优于各自回归模型中任何单一变量。结论:本研究所建立的多变量预测模型能更早期、更有效地预测PUL患者的妊娠状态,对早期PUL患者妊娠结局的预测有较大的临床价值。     

Quality and Presence of Behaviour Change Techniques in Mobile Apps for the Mediterranean Diet: A Content Analysis of Android Google Play and Apple App Store Apps

Smartphone apps might represent an opportunity to promote adherence to the Mediterranean diet (MedDiet). This study aimed to evaluate the quality of commercially available apps for the MedDiet and the presence of behavioural change techniques (BCTs) used by these apps. A systematic search was conducted on the Apple App and Google Play stores in November 2021. Apps were included if they provided information on the MedDiet or if their objective was to promote a healthy lifestyle through adherence to the MedDiet. Eligible apps were independently evaluated by two reviewers with regard to their quality (engagement, functionality, aesthetics and information quality) using the 5-point Mobile App Rating Scale (MARS; with higher scores indicating higher quality), and the presence of BCTs using an established 26-item BCT taxonomy. Of the 55 analysed apps, 52 (94.5%) were free, 50 (90.9%) provided recipe ideas, 29 (52.7%) provided meal plans, and 22 (40%) provided information on the health benefits of the MedDiet. The overall quality mean MARS score was 2.84 (standard deviation (SD) = 0.42), with functionality being the highest scored MARS domain (mean = 3.58, SD = 0.44) and engagement the lowest (mean = 2.29, SD = 0.61). The average number of BCTs in the analysed apps was 2.3 (SD = 1.4; range: 0–6 per app). The number of BCTs was positively correlated with app information quality (r_rho = 0.269, p = 0.047), overall MARS score (r_rho = 0.267, p = 0.049), app subjective quality (r_rho = 0.326, p = 0.015) and app-specific quality (r_rho = 0.351, p = 0.009). These findings suggest that currently available apps might provide information on the MedDiet, but the incorporation of more BCTs is warranted to maximise the potential for behaviour change towards the MedDiet.     

Applying Machine Learning to Determine 25(OH)D Threshold Levels Using Data from the AMATERASU Vitamin D Supplementation Trial in Patients with Digestive Tract Cancer

Some controversy remains on thresholds for deficiency or sufficiency of serum 25-hydroxyvitamin D (25(OH)D) levels. Moreover, 25(OH)D levels sufficient for bone health might differ from those required for cancer survival. This study aimed to explore these 25(OH)D threshold levels by applying the machine learning method of multivariable adaptive regression splines (MARS) in post hoc analyses using data from the AMATERASU trial, which randomly assigned Japanese patients with digestive tract cancer to receive vitamin D or placebo supplementation. Using MARS, threshold 25(OH)D levels were estimated as 17 ng/mL for calcium and 29 ng/mL for parathyroid hormone (PTH). Vitamin D supplementation increased calcium levels in patients with baseline 25(OH)D levels ≤17 ng/mL, suggesting deficiency for bone health, but not in those >17 ng/mL. Vitamin D supplementation improved 5-year relapse-free survival (RFS) compared with placebo in patients with intermediate 25(OH)D levels (18–28 ng/mL): vitamin D, 84% vs. placebo, 71%; hazard ratio, 0.49; 95% confidence interval, 0.25–0.96; p = 0.04. In contrast, vitamin D supplementation did not improve 5-year RFS among patients with low (≤17 ng/mL) or with high (≥29 ng/mL) 25(OH)D levels. MARS might be a reliable method with the potential to eliminate guesswork in the estimation of threshold values of biomarkers.     

武汉市116例境外输入性恶性疟临床治疗效果分析

目的分析和评价境外输入性恶性疟患者临床特点及青蒿素类药物治疗效果,促进合理规范用药,提高临床治愈率。方法收集武汉市省级疟疾定点医院收治的116例输入性恶性疟病例,对患者临床症状及体征、青蒿素药物使用情况及用药效果等进行回顾性分析。结果输入性恶性疟临床表现复杂多样且并发症较多,主要为发热、纳差、贫血、黄疸等;使用青蒿素治疗后症状均减轻或消失,无明显药物不良反应,退热中位数为2d,疟原虫转阴中位数为3d,临床症状缓解中位数为3d;109例普通型恶性疟采用青蒿素复方口服或青蒿琥酯针剂联合青蒿素复方口服治疗,其中30例口服青蒿素复方治疗1~4d, 1例复燃;79例青蒿琥酯针剂联合青蒿素复方口服治疗4~6d, 4例出现复燃。7例重症病例采用青蒿琥酯针剂联合口服多种青蒿素复方长疗程大剂量后均治愈,总治愈率95.69%(111/116)。结论患者以口服和注射青蒿素杀灭疟原虫治疗恶性疟安全有效,治愈率高,但部分患者停用青蒿素类药物后出现疟原虫复燃,为遏制患者重症和抗药风险,应"规范、全程、足量"使用青蒿素类药品。     

Effects of Artemisinin on Photosystem II Performance of Microcystis aeruginosa by In Vivo Chlorophyll Fluorescence

Effects of artemisinin (derived from Artemisia annua) on the photosynthetic activity of Microcystis aeruginosa was investigated by using chlorophyll a (Chl a) fluorescence transient O-J-I–P and JIP-test after exposure to elevated artemisinin concentration. High artemisinin concentration resulted in a significant suppression in photosynthesis and respiration. Results showed that the OJIP curves flattened and the maximal fluorescence yield reached at the J step under artemisinin stress. The decreased values of the energy needed for the RCs’ closure (Sm) and the number of oxidation and reduction (N) suggested that the reduction times of primary bound plastoquinone (Q_A) was also decreased. The absorption flux (ABS/RC) per photosystem II (PSII) reaction center and the electron transport flux (ET₀/RC) decreased with increasing artemisinin concentration. Excess artemisinin had little effect on the trapping flux (TR₀/RC). The results showed that the decrease of photosynthesis in exposure to excess artemisinin may be a result of the inactivation of PSII reaction centers and the inhibition of electron transport in the acceptor side.     

German Mobile Apps for Patients with Psoriatic Arthritis: Systematic App Search and Content Analysis

BackgroundPsoriatic arthritis (PsA) is a chronic musculoskeletal disease. Functional limitations result in significant patient burden. The use of smartphone apps to complement medical treatments could increase patients’ adherence resulting in better clinical outcome and consequently increased quality of life.ObjectiveThis study aimed to assess and compare currently available smartphone apps for PsA patients using a standardized rating scale.MethodsWe conducted a systematic screening of apps for PsA patients in German App Stores and selected apps meeting inclusion criteria: availability in both app stores, German or English language, relevant for patients with psoriatic arthritis, not developed for congress use or clinical studies, no fee-based apps. Three reviewers evaluated apps using the Mobile Application Rating Scale (MARS). Analysis included technical features and mean scores as well as interrater agreement and linear regression. The best-rated app was then rated by PsA patients.ResultsAmong 232 screened apps, 8 met all inclusion criteria. The app “Rheuma-Auszeit” scored highest with a healthcare provider MARS rating of 4.4 and a patient userMARS rating of 3.7. Subjective quality was lower for all apps. App store star ratings and MARS ratings were not significantly correlated (P=0.34). No app was Conformité Européenne (CE) certified and no underlying clinical studies were found.ConclusionsApp quality was heterogenous and generally low. Star ratings – the only accessible tool for patients – does not suffice to identify high-quality apps. We suggest involving patients in development of apps. Since patients and physicians have different perceptions of app quality, we recommend a joint development and evaluation process.     

The Benefits of Telephone-Access Medical Consultation

A major disadvantage of rural medical practice is the limited reserve of consultative options. To determine the perceived clinical utility and educational impact of the West Virginia University Medical Access and Referral System (MARS), a 24-hour prompt telephone-consultation service, a mailed questionnaire was administered to 303 West Virginia clinicians who had used MARS for infectious disease problems. The overall questionnaire response rate was 62 percent. Callers included family practitioners (35%), medical specialists (32%), surgical specialists (13%), pediatricians (11%), obstetricians (5%), and nonphysicians (4%). Major referral questions posed were therapeutic (60%), diagnostic (48%), and epidemiologic (10%) in nature. On a scale of 1 (not useful) to 5 (very useful), survey responders rated the overall clinical usefulness of MARS as either a 4 (22%) or 5 (76%). Callers felt that MARS consultation assisted in accurate case diagnosis in 80 percent of cases, and aided in successful therapeutic management of 96 percent of cases. An educational benefit was reported by 96 percent of responders. Physicians located in more rural, underserved areas tended to use MARS to a greater degree than colleagues in more populated, medically accessible areas (P < 0.005). These findings suggest that an academic telephone-access consultation program can be a clinically relevant and educational consultative tool for practicing clinicians, especially those located in rural areas.     

多酚类化合物抗氧化活性的电拓扑模型

目的建立多酚类化合物抗氧化活性和原子类型电性拓扑指数的定量构效关系模型,探讨影响其结构的因素。方法采用拓扑指数法计算多酚类分子的结构参数,运用最佳变量子集回归法初步建立它们4种抗氧化活性的QSAR模型。结果所建4个QSAR模型均通过Fischer和VIF检验,具有良好的稳健性。结论分子内=O、=C<、—CH3、—COO—等基团对多酚类化合物的抗氧化活性具有很大影响。     

In vitro activity of artemisinin, its derivatives, and pyronaridine against different strains of Plasmodium falciparum

The activities of artemisinin (qinghaosu), dihydroartemisinin (dihydroqinghaosu), artemether and pyronaridine were tested in a 48 h in vitro assay against 3 chloroquine-sensitive and one chloroquine-resistant strains of Plasmodium falciparum. Growth inhibition was modelled with a sigmoid Emax model. All compounds markedly inhibited the growth of all strains although, for the chloroquine-resistant strain, merozoites were detected at concentrations as high as 10(-4) M of artemisinin, dihydroartemisinin and artemether. Dihydroartemisinin, artemether and pyronaridine appeared to be more potent than artemisinin, with EC50 values of 4.7-23 nM, 0.98-6.1 nM and 4.4-18 nM respectively, while the EC50 value of artemisinin was 3-108 nM against all 4 strains.     

Prediction of aerobic fitness without stress testing and applicability to epidemiological studies: a systematic review

Cardiorespiratory fitness is used as an independent factor for evaluating risk of all-cause mortality, but mainly from coronary heart disease. Nevertheless, evaluation of fitness based on stress tests poses numerous epidemiological difficulties. Alternative forms of evaluation have therefore been suggested using non-exercise-based regression models. This study aimed to analyze these models and their applicability to epidemiological studies. A systematic review was conducted of articles published from 1966 to 2002. The models were classified according to: (a) theoretical justification for the explanatory variables included in the model; (b) validation criteria (gold standard); (c) regression models fully reported, including standard error of estimation; and (d) cross-validation. The search process yielded 23 studies, five of which met all established quality criteria. The authors conclude that cardiorespiratory evaluation by non-exercise-based models could be feasible in epidemiological studies. However, few equations appear to meet the minimum external validation requirements to provide data that could be generalized to large populations.     

Dosimetry in MARS spectral CT: TOPAS Monte Carlo simulations and ion chamber measurements

Spectral computed tomography (CT) is an up and coming imaging modality which shows great promise in revealing unique diagnostic information. Because this imaging modality is based on X-ray CT, it is of utmost importance to study the radiation dose aspects of its use. This study reports on the implementation and evaluation of a Monte Carlo simulation tool using TOPAS for estimating dose in a pre-clinical spectral CT scanner known as the MARS scanner. Simulated estimates were compared with measurements from an ionization chamber. For a typical MARS scan, TOPAS estimated for a 30 mm diameter cylindrical phantom a CT dose index (CTDI) of 29.7 mGy; CTDI was measured by ion chamber to within 3% of TOPAS estimates. Although further development is required, our investigation of TOPAS for estimating MARS scan dosimetry has shown its potential for further study of spectral scanning protocols and dose to scanned objects.     

Multivariate Adaptative Regression Splines (MARS), una alternativa para el análisis de series de tiempo

Multivariate Adaptive Regression Splines (MARS) is a non-parametric modelling method that extends the linear model, incorporating nonlinearities and interactions between variables. It is a flexible tool that automates the construction of predictive models: selecting relevant variables, transforming the predictor variables, processing missing values and preventing overshooting using a self-test. It is also able to predict, taking into account structural factors that might influence the outcome variable, thereby generating hypothetical models. The end result could identify relevant cut-off points in data series. It is rarely used in health, so it is proposed as a tool for the evaluation of relevant public health indicators. For demonstrative purposes, data series regarding the mortality of children under 5 years of age in Costa Rica were used, comprising the period 1978–2008.