Detecting potential outliers in longitudinal data with time-dependent covariates

BACKGROUND

Outliers can influence regression model parameters and change the direction of the estimated effect, over-estimating or under-estimating the strength of the association between a response variable and an exposure of interest. Identifying visit-level outliers from longitudinal data with continuous time-dependent covariates is important when the distribution of such variable is highly skewed.

OBJECTIVES

The primary objective was to identify potential outliers at follow-up visits using interquartile range (IQR) statistic and assess their influence on estimated Cox regression parameters.

METHODS

Study was motivated by a large TEDDY dietary longitudinal and time-to-event data with a continuous time-varying vitamin B12 intake as the exposure of interest and development of Islet Autoimmunity (IA) as the response variable. An IQR algorithm was applied to the TEDDY dataset to detect potential outliers at each visit. To assess the impact of detected outliers, data were analyzed using the extended time-dependent Cox model with robust sandwich estimator. Partial residual diagnostic plots were examined for highly influential outliers.

RESULTS

Extreme vitamin B12 observations that were cases of IA had a stronger influence on the Cox regression model than non-cases. Identified outliers changed the direction of hazard ratios, standard errors, or the strength of association with the risk of developing IA.

CONCLUSION

At the exploratory data analysis stage, the IQR algorithm can be used as a data quality control tool to identify potential outliers at the visit level, which can be further investigated.

Comparison of Rally Length between Women and Men in High-Level Spanish Volleyball

The aim of this study was to evaluate whether the rally length in high-level Spanish volleyball was longer in women than in men. A total of 1,786 rallies were observed: 792 for women and 994 for men. The recorded variables were match (quarter-final 1, quarter-final 2, semi-final 1, semi-final 2, final), gender (men, women), rally length (seconds), pseudo-rally (no, yes), and terminal event (ball out of sight, ball in/out, fault). Different non-parametric statistical techniques were used to compare the rally length between groups or subsets of data, i.e., the Kruskal-Wallis H test, the Mann-Whitney U test, quantile regression, and survival analysis. The mean and median rally length was significantly and slightly longer in women than in men. The rally length difference between genders was barely 1 s in quantile 0.5 or median, while in quantile 0.95, it was just over 4 s. In women, the probability of ending the rally at 3.9, 5.1, 10.2, and 43.9 s (at 4.4, 6.3, 11.6, and 43.9 s without pseudo-rallies) was 25%, 50%, 75%, and 100%, respectively. In men, the probability of ending the rally at 3.2, 4.3, 7.9, and 29.1 s (at 3.9, 4.8, 8.8, and 29.1 s without pseudo-rallies) was 25%, 50%, 75%, and 100%, respectively. These temporal thresholds can help volleyball coaches to train their players in a coherent manner.

Win ratio approach for analyzing composite time-to-event endpoint with opposite treatment effects in its components

There is an increasing interest in the use of win ratio with composite time-to-event due to its flexibility in combining component endpoints. Exploring this flexibility further, one interesting question is in assessing the impact when there is a difference in treatment effect in the component endpoints. For example, the active treatment may prolong the time to occurrence of the negative event such as death or ventilation; meanwhile, the treatment effect may also shorten the time to achieving positive events, such as recovery or improvement. Notably, this portrays a situation where the treatment effect on time to recovery is in a different direction of benefit compared to the time to ventilation or death. Under such circumstances, if a single endpoint is used, the benefit gained for other individual outcomes is not counted and is diminished. As consequence, the study may need a larger sample size to detect a significant effect of treatment. Such a scenario can be handled by win ratio in a novel way by ranking component events, which is different from the usual composite endpoint approach such as time-to-first event. To evaluate how the different directions of treatment effect on component endpoints will impact the win ratio analysis, we use a Clayton copula-based bivariate survival simulation to investigate the correlation of component time-to-event. Through simulation, we found that compared to the marginal model using single endpoints, the win ratio analysis on composite endpoint performs better, especially when the correlation between two events is weak. Then, we applied the methodology to an infectious disease progression simulated study motivated by COVID-19. The application demonstrates that the win ratio approach offers advantages in empirical power compared to the traditional Cox proportional hazard approach when there is a difference in treatment effect in the marginal events.

An open-source toolbox for Multi-patient Intracranial EEG Analysis (MIA)

Intracranial EEG (iEEG) performed during the pre-surgical evaluation of refractory epilepsy provides a great opportunity to investigate the neurophysiology of human cognitive functions with exceptional spatial and temporal precisions. A difficulty of the iEEG approach for cognitive neuroscience, however, is the potential variability across patients in the anatomical location of implantations and in the functional responses therein recorded. In this context, we designed, implemented, and tested a user-friendly and efficient open-source toolbox for Multi-Patient Intracranial data Analysis (MIA), which can be used as standalone program or as a Brainstorm plugin. MIA helps analyzing event related iEEG signals while following good scientific practice recommendations, such as building reproducible analysis pipelines and applying robust statistics. The signals can be analyzed in the temporal and time-frequency domains, and the similarity of time courses across patients or contacts can be assessed within anatomical regions. MIA allows visualizing all these results in a variety of formats at every step of the analysis. Here, we present the toolbox architecture and illustrate the different steps and features of the analysis pipeline using a group dataset collected during a language task.

Functional Instability of the Second to Fifth Metacarpophalangeal Joints

BACKGROUND

Clinically, functional instability (FI) of metacarpophalangeal joints (MCPJ) is not considered to represent a pathology. This excessive mobility can be detected by the application of external forces to a MCPJ at different angles. Our study aimed to measure the FI of 2nd to 5th MCPJ.

MATERIALS AND METHODS

A group of 36 healthy right-handed individuals were enrolled. The value of FI was measured in millimeters and verified by a CT scan. Statistical calculations was made in Statistica v.10.0.

RESULTS

The largest values of the right and left-hand finger posterior displacement (FI) in the second to fifth MCPJ were obtained in the neutral position 0° (p<0.05). Measurements of volar displacement of the proximal phalanx second to fifth MCPJs in both hands revealed higher values at position 0° (p<0.05).

CONCLUSIONS

1. FI of the second to fifth MCPJs is determined by anatomical configuration and dynamic stabilizers. 2. Instability measurements show that posterior displacement of the proximal phalanges is greater by at least 1 mm in the 0° neutral position.

Non-parametric treatment time-lag effect estimation

In general, the change point problem considers inference of a change in distribution for a set of time-ordered observations. This has applications in a large variety of fields, and can also apply to survival data. In survival analysis, most existing methods compare two treatment groups for the entirety of the study period. Some treatments may take a length of time to show effects in subjects. This has been called the time-lag effect in the literature, and in cases where time-lag effect is considerable, such methods may not be appropriate to detect significant differences between two groups. In this paper, we propose a novel non-parametric approach for estimating the point of treatment time-lag effect by using an empirical divergence measure. Theoretical properties of the estimator are studied. The results from the simulated data and the applications to real data examples support our proposed method.

The asymptotic distribution of the Net Benefit estimator in presence of right-censoring

The benefit-risk balance is a critical information when evaluating a new treatment. The Net Benefit has been proposed as a metric for the benefit-risk assessment, and applied in oncology to simultaneously consider gains in survival and possible side effects of chemotherapies. With complete data, one can construct a U-statistic estimator for the Net Benefit and obtain its asymptotic distribution using standard results of the U-statistic theory. However, real data is often subject to right-censoring, e.g. patient drop-out in clinical trials. It is then possible to estimate the Net Benefit using a modified U-statistic, which involves the survival time. The latter can be seen as a nuisance parameter affecting the asymptotic distribution of the Net Benefit estimator. We present here how existing asymptotic results on U-statistics can be applied to estimate the distribution of the net benefit estimator, and assess their validity in finite samples. The methodology generalizes to other statistics obtained using generalized pairwise comparisons, such as the win ratio. It is implemented in the R package BuyseTest (version 2.3.0 and later) available on Comprehensive R Archive Network.

HyPyP: a Hyperscanning Python Pipeline for inter-brain connectivity analysis

The bulk of social neuroscience takes a 'stimulus-brain' approach, typically comparing brain responses to different types of social stimuli, but most of the time in the absence of direct social interaction. Over the last two decades, a growing number of researchers have adopted a 'brain-to-brain' approach, exploring similarities between brain patterns across participants as a novel way to gain insight into the social brain. This methodological shift has facilitated the introduction of naturalistic social stimuli into the study design (e.g. movies) and, crucially, has spurred the development of new tools to directly study social interaction, both in controlled experimental settings and in more ecologically valid environments. Specifically, 'hyperscanning' setups, which allow the simultaneous recording of brain activity from two or more individuals during social tasks, has gained popularity in recent years. However, currently, there is no agreed-upon approach to carry out such 'inter-brain connectivity analysis', resulting in a scattered landscape of analysis techniques. To accommodate a growing demand to standardize analysis approaches in this fast-growing research field, we have developed Hyperscanning Python Pipeline, a comprehensive and easy open-source software package that allows (social) neuroscientists to carry-out and to interpret inter-brain connectivity analyses.

Intra- and inter-rater reliability of Fugl-Meyer Assessment of Upper Extremity in stroke

OBJECTIVE

The Fugl-Meyer Assessment of Upper Extremity (FMA-UE) is recommended for evaluation of sensorimotor impairment post stroke, but the item-level reliability of the scale is unknown. The study aims to determine intra- and inter-rater reliability of the FMA-UE at item-, subscale- and total score level in patients with early subacute stroke.

DESIGN

Intra/inter-rater reliability.

SUBJECTS

Sixty consecutively included patients with stroke (mean age 65.9 years) admitted to Central Military Hospital of Colombia, Bogota.

METHODS

Two physiotherapists scored FMA-UE independently on 2 consecutive days within 10 days post stroke. A rank-based statistical method for paired ordinal data was used to assess the level of agreement, systematic and random disagreements.

RESULTS

Systematic disagreements either in position or concentration were detected in 4 items of the shoulder section. The item level intra- and inter-rater agreement was high (79100%). The 70% agreement was also reached for the subscales and the total score when 13-point difference was accepted.

CONCLUSION

The FMA-UE is reliable both within and between raters in patients with stroke in the early subacute phase. A wider international use of FMA-UE will allow comparison of stroke recovery between regions and countries and thereby potentially improve the quality of care and rehabilitation in persons with stroke worldwide.

Clustering the Brain With "CluB": A New Toolbox for Quantitative Meta-Analysis of Neuroimaging Data

In this paper we describe and validate a new coordinate-based method for meta-analysis of neuroimaging data based on an optimized hierarchical clustering algorithm: CluB (Clustering the Brain). The CluB toolbox permits both to extract a set of spatially coherent clusters of activations from a database of stereotactic coordinates, and to explore each single cluster of activation for its composition according to the cognitive dimensions of interest. This last step, called “cluster composition analysis,” permits to explore neurocognitive effects by adopting a factorial-design logic and by testing the working hypotheses using either asymptotic tests, or exact tests either in a classic inference, or in a Bayesian-like context. To perform our validation study, we selected the fMRI data from 24 normal controls involved in a reading task. We run a standard random-effects second level group analysis to obtain a “Gold Standard” of reference. In a second step, the subject-specific reading effects (i.e., the linear t-contrast “reading > baseline”) were extracted to obtain a coordinates-based database that was used to run a meta-analysis using both CluB and the popular Activation Likelihood Estimation method implemented in the software GingerALE. The results of the two meta-analyses were compared against the “Gold Standard” to compute performance measures, i.e., sensitivity, specificity, and accuracy. The GingerALE method obtained a high level of accuracy (0.967) associated with a high sensitivity (0.728) and specificity (0.971). The CluB method obtained a similar level of accuracy (0.956) and specificity (0.969), notwithstanding a lower level of sensitivity (0.14) due to the lack of prior Gaussian transformation of the data. Finally, the two methods obtained a good-level of concordance (AC₁ = 0.93). These results suggested that methods based on hierarchical clustering (and post-hoc statistics) and methods requiring prior Gaussian transformation of the data can be used as complementary tools, with the GingerALE method being optimal for neurofunctional mapping of pooled data according to simpler designs, and the CluB method being preferable to test more specific, and localized, neurocognitive hypotheses according to factorial designs.

Social media and bitcoin metrics: which words matter

We develop a new Data-Driven Phasic Word Identification (DDPWI) methodology to determine which words matter as the bitcoin pricing dynamic changes from one phase to another. With Google search volumes as a baseline, we find that Reddit submissions are both correlated with Google and have a comparable relationship with a variety of bitcoin metrics, using Spearman's rho. Reddit provides complete access to the text of submissions. Rather than associating sentiment with market activity, we describe the DDPWI method for finding specific 'price dynamic' words associated with changes in the bitcoin pricing pattern through 2017 and 2018. We assess the significance of these changes using Wilcoxon Rank-Sum Tests with Bonferroni corrections. These price dynamic words are used to pull out associated words in the submissions thereby providing the context to their use. For example, the price dynamic word 'ban', which became significantly higher in frequency as prices fell, occurred in the context of both government regulation and internet companies banning cryptocurrency adverts. This approach could be used more generally to look at social media and discussion forums at a granular level identifying specific words that impact the metric under investigation rather than overall sentiment.

A Tool for Brain-Wide Quantitative Analysis of Molecular Data upon Projection into a Planar View of Choice

Several techniques, allowing the reconstruction and visualization of functional, anatomical or molecular information from tissue and organ slices, have been developed over the years. Yet none allow direct comparison without reprocessing the same slices. Alternative methods using publicly available reference maps like the Allen Brain Atlas lack flexibility with respect to age and species. We propose a new approach to reconstruct a segmented region of interest from serial slices by projecting the optical density values representing a given molecular signal to a plane of view of choice, and to generalize the results into a reference map, which is built from the individual maps of all animals under study. Furthermore, to allow quantitative comparison between experimental conditions, a non-parametric pseudo t-test has been implemented. This new mapping tool was applied, optimized and validated making use of an in situ hybridization dataset that represents the spatiotemporal expression changes for the neuronal activity reporter gene zif268, in relation to cortical plasticity induced by monocular enucleation, covering the entire mouse visual cortex. The created top view maps of the mouse brain allow precisely delineating and interpreting 11 extrastriate areas surrounding mouse V1. As such, and because of the opportunity to create a planar projection of choice, these molecular maps can in the future easily be compared with functional or physiological imaging maps created with other techniques such as Ca2+, flavoprotein and optical imaging.

When the Single Matters more than the Group (II): Addressing the Problem of High False Positive Rates in Single Case Voxel Based Morphometry Using Non-parametric Statistics

In recent years, an increasing number of studies have used Voxel Based Morphometry (VBM) to compare a single patient with a psychiatric or neurological condition of interest against a group of healthy controls. However, the validity of this approach critically relies on the assumption that the single patient is drawn from a hypothetical population with a normal distribution and variance equal to that of the control group. In a previous investigation, we demonstrated that family-wise false positive error rate (i.e., the proportion of statistical comparisons yielding at least one false positive) in single case VBM are much higher than expected (Scarpazza et al., 2013). Here, we examine whether the use of non-parametric statistics, which does not rely on the assumptions of normal distribution and equal variance, would enable the investigation of single subjects with good control of false positive risk. We empirically estimated false positive rates (FPRs) in single case non-parametric VBM, by performing 400 statistical comparisons between a single disease-free individual and a group of 100 disease-free controls. The impact of smoothing (4, 8, and 12 mm) and type of pre-processing (Modulated, Unmodulated) was also examined, as these factors have been found to influence FPRs in previous investigations using parametric statistics. The 400 statistical comparisons were repeated using two independent, freely available data sets in order to maximize the generalizability of the results. We found that the family-wise error rate was 5% for increases and 3.6% for decreases in one data set; and 5.6% for increases and 6.3% for decreases in the other data set (5% nominal). Further, these results were not dependent on the level of smoothing and modulation. Therefore, the present study provides empirical evidence that single case VBM studies with non-parametric statistics are not susceptible to high false positive rates. The critical implication of this finding is that VBM can be used to characterize neuroanatomical alterations in individual subjects as long as non-parametric statistics are employed.

Classification and regression tree analysis vs. multivariable linear and logistic regression methods as statistical tools for studying haemophilia

INTRODUCTION

Haemophilia is a rare genetic haemorrhagic disease characterized by partial or complete deficiency of coagulation factor VIII, for haemophilia A, or IX, for haemophilia B. As in any other medical research domain, the field of haemophilia research is increasingly concerned with finding factors associated with binary or continuous outcomes through multivariable models. Traditional models include multiple logistic regressions, for binary outcomes, and multiple linear regressions for continuous outcomes. Yet these regression models are at times difficult to implement, especially for non-statisticians, and can be difficult to interpret.

AIMS

The present paper sought to didactically explain how, why, and when to use classification and regression tree (CART) analysis for haemophilia research.

MATERIALS & METHODS

The CART method is non-parametric and non-linear, based on the repeated partitioning of a sample into subgroups based on a certain criterion. Breiman developed this method in 1984. Classification trees (CTs) are used to analyse categorical outcomes and regression trees (RTs) to analyse continuous ones.

RESULTS

The CART methodology has become increasingly popular in the medical field, yet only a few examples of studies using this methodology specifically in haemophilia have to date been published. Two examples using CART analysis and previously published in this field are didactically explained in details.

CONCLUSION

There is increasing interest in using CART analysis in the health domain, primarily due to its ease of implementation, use, and interpretation, thus facilitating medical decision-making. This method should be promoted for analysing continuous or categorical outcomes in haemophilia, when applicable.

First Report on Self-Rated Health in a Nationally-Representative Sample of Iranian Adolescents: The CASPIAN-iii study

OBJECTIVE

To evaluate predictive factors of adolescents' appraisal of their health.

METHODS

The nationwide study, entitled "Childhood and Adolescence Surveillance and Prevention of Adult Non-communicable Diseases (CASPIAN) study", was conducted in 2010 among Iranian school students, aged 10-18. In addition to demographic factors and physical examination, variables as family structure, nutrition habits, physical activity, smoking, hygienic habits, violence, school attachment, family smoking, and family history of chronic diseases were assessed. The dependent variable is the self-rated health (SRH) and it was measured by 12 items, which had already been combined through latent class analysis. We had taken a dichotomous variable, i.e. the higher values indicate better SRH. The dependent variable was regressed on all predictors by generalized additive models.

RESULTS

75% of adolescents had a good SRH. The linear and smooth effects of independent variables on SRH were observed. Among all the variables, physical activity had a positive linear effect on SRH (β = 0.08, P value = 0.003). Smoking, violence, and family history of disease associated to SRH non-linearly (P value < 0.05). Family smoking (β = -0.01) and hygienic habits (β = 0.27) related to SRH both linearly and non-linearly.

CONCLUSIONS

Physical health and high risk behavior, either of linear or non-linear effect, are factors, which seem to shape the adolescents' perception of health.

Non-parametric statistical thresholding for sparse magnetoencephalography source reconstructions

Uncovering brain activity from magnetoencephalography (MEG) data requires solving an ill-posed inverse problem, greatly confounded by noise, interference, and correlated sources. Sparse reconstruction algorithms, such as Champagne, show great promise in that they provide focal brain activations robust to these confounds. In this paper, we address the technical considerations of statistically thresholding brain images obtained from sparse reconstruction algorithms. The source power distribution of sparse algorithms makes this class of algorithms ill-suited to "conventional" techniques. We propose two non-parametric resampling methods hypothesized to be compatible with sparse algorithms. The first adapts the maximal statistic procedure to sparse reconstruction results and the second departs from the maximal statistic, putting forth a less stringent procedure that protects against spurious peaks. Simulated MEG data and three real data sets are utilized to demonstrate the efficacy of the proposed methods. Two sparse algorithms, Champagne and generalized minimum-current estimation (G-MCE), are compared to two non-sparse algorithms, a variant of minimum-norm estimation, sLORETA, and an adaptive beamformer. The results, in general, demonstrate that the already sparse images obtained from Champagne and G-MCE are further thresholded by both proposed statistical thresholding procedures. While non-sparse algorithms are thresholded by the maximal statistic procedure, they are not made sparse. The work presented here is one of the first attempts to address the problem of statistically thresholding sparse reconstructions, and aims to improve upon this already advantageous and powerful class of algorithm.

Gender and condylar differences in distal femur morphometry clarified by automated computer analyses

We elucidated the gender and condylar effects on distal femur morphology (DFM) while evaluating a newly developed computational framework that enables fully automated analyses of DFM in an objectively defined sagittal plane. Ninety high-resolution CT-acquired distal femur models from 51 males and 39 females were analyzed. The models were accurately characterized (mean least-squares fitting residual <0.16 mm), and re-oriented to a unified sagittal plane; three morphometric measures were extracted from each model: the semi-major (a) and semi-minor (b) axis lengths of the best-fitted ellipse, and the radius (r) of the smallest flexion facet-a circle with the smallest radius best-fitted to the posterior articulating surface. Statistical analyses employing nonparametric repeated-measures ANOVA found: no significance difference between condyles or between limbs in any of the morphometric measures; significant gender effects on a, b, and r, but no gender effect on the aspect ratio (a/b). An inspection of statistical distributions of medial-lateral condyle size differences also revealed a gender difference. The findings promote a better understanding of DFM and its relation to knee mechanics and have implications on computer-aided surgery of the knee and gender-specific implant design.