OROlivier Renaud

Olivier Renaud

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Olivier Renaud has a total of 148 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Widespread use of invalid statistical tests in biomedical machine learning

    by Tianchu Zeng, Hetu Li, Shaoshi Zhang, Yan Quan Tan, Fang Tian, Csaba Orban, Lijun An, Wanyu Che, Jingwen Cheng, Joanna Su Xian Chong, et al.

    Machine learning is accelerating biomedical research. Cross-validation is widely used to compare predictive performance — not only to benchmark algorithms, but also to inform scientific applications, such as ranking biomarkers. However, prediction performance estimates across cross-validation folds are not independent. Standard tests for comparing prediction performance (e.g., paired t-test) assume independence and can therefore inflate false positive rates. In a PRISMA-guided meta-analysis of 210 studies (impact factor ≥15, 1 June 2020 – 1 June 2025), we find that 97% ignored fold dependence when comparing prediction performance. This problem is ubiquitous across scientific fields and unaffected by impact factor, rigor-promoting policies, or open science practices. Simulations across 420 scenarios spanning four diverse datasets show that ignoring fold dependence leads to invalid false positive control in most settings. Repeated cross-validation further compounds this problem, with false positive rates rising toward 100% as the number of repetitions grows. Existing fold-dependence-aware tests rely on strong assumptions because the variance of fold-level statistics and the between-fold correlation cannot be disentangled under standard cross-validation. We therefore propose the SHARP (Split-HAlf RePeated) test, a simple modification to standard cross-validation that enables direct estimation of variance and correlation. Benchmarked against 12 tests, SHARP provides the best overall balance of false-positive control, statistical power, and confidence-interval calibration across simulation schemes. We conclude by providing best practices and reporting guidelines for valid model comparison inference in biomedical machine learning and beyond.


  • Open Access

    The effect of d-cycloserine on brain processing of breathlessness over pulmonary rehabilitation: an experimental medicine study

    by Sarah L. Finnegan, Olivia K. Harrison, Sara Booth, Andrea Dennis, Martyn Ezra, Catherine J. Harmer, Mari Herigstad, Bryan Guillaume, Thomas E. Nichols, Najib M. Rahman, et al.

    Research question: Pulmonary rehabilitation is the best treatment for chronic breathlessness in COPD but there remains an unmet need to improve efficacy. Pulmonary rehabilitation has strong parallels with exposure-based cognitive behavioural therapies (CBT), both clinically and in terms of brain activity patterns. The partial N-methyl-d-aspartate (NMDA)-receptor agonistd-cycloserine has shown promising results in enhancing efficacy of CBT, thus we hypothesised that it would similarly augment the effects of pulmonary rehabilitation in the brain. Positive findings would support further development in phase 3 clinical trials.Methods: 72 participants with mild-to-moderate COPD were recruited to a double-blind pre-registered (ClinicalTrials.govidentifier:NCT01985750) experimental medicine study running parallel to a pulmonary rehabilitation course. Participants were randomised to 250 mgd-cycloserine or placebo, administered immediately prior to the first four sessions of pulmonary rehabilitation. Primary outcome measures were differences betweend-cycloserine and placebo in brain activity in the anterior insula, posterior insula, anterior cingulate cortices, amygdala and hippocampus following completion of pulmonary rehabilitation. Secondary outcomes included the same measures at an intermediate time point and voxel-wise difference across wider brain regions. An exploratory analysis determined the interaction with breathlessness anxiety.Results: No difference betweend-cycloserine and placebo groups was observed across the primary or secondary outcome measures.d-cycloserine was shown instead to interact with changes in breathlessness anxiety to dampen reactivity to breathlessness cues. Questionnaire and measures of respiratory function showed no group difference. This is the first study testing brain-active drugs in pulmonary rehabilitation. Rigorous trial methodology and validated surrogate end-points maximised statistical power.Conclusion: Although increasing evidence supports therapeutic modulation of NMDA pathways to treat symptoms, we conclude that a phase 3 clinical trial ofd-cycloserine would not be worthwhile.


  • Open Access

    The Cluster Depth Tests: Toward Point-Wise Strong Control of the Family-Wise Error Rate in Massively Univariate Tests with Application to M/EEG

    by Jaromil Frossard, Olivier Renaud

    The cluster mass test has been widely used for massively univariate tests in M/EEG, fMRI and, recently, pupillometry analysis. It is a powerful method for detecting effects while controlling weakly the family-wise error rate (FWER), although its correct interpretation can only be performed at the cluster level without any point-wise conclusion. It implies that the discoveries of a cluster mass test cannot be precisely localized in time or in space. We propose a new multiple comparison procedure, the cluster depth tests, that both controls the FWER while allowing an interpretation at the time point level. The simulation study shows that the cluster depth tests achieve large power and guarantee the FWER even in the presence of physiologically plausible effects. By having an interpretation at the time point/voxel level, the cluster depth tests make it possible to take full advantage of the high temporal resolution of EEG recording to precisely time the appearance of an effect.

  • Open Access

    Shared and Anxiety-Specific Pediatric Psychopathology Dimensions Manifest Distributed Neural Correlates

    by Julia O. Linke, Rany Abend, Katharina Kircanski, Michal Clayton, Caitlin Stavish, Brenda E. Benson, Melissa A. Brotman, Olivier Renaud, Stephen M. Smith, Thomas E. Nichols, et al.


  • Open Access

    Permutation inference for canonical correlation analysis

    by Anderson M. Winkler, Olivier Renaud, Stephen M. Smith, Thomas E. Nichols


  • Open Access

    Permutation Inference for Canonical Correlation Analysis

    by Anderson M. Winkler, Olivier Renaud, Stephen M. Smith, Thomas E. Nichols

    Canonical correlation analysis (CCA) has become a key tool for population neuroimaging, allowing investigation of associations between many imaging and non-imaging measurements. As other variables are often a source of variability not of direct interest, previous work has used CCA on residuals from a model that removes these effects, then proceeded directly to permutation inference. We show that such a simple permutation test leads to inflated error rates. The reason is that residualisation introduces dependencies among the observations that violate the exchangeability assumption. Even in the absence of nuisance variables, however, a simple permutation test for CCA also leads to excess error rates for all canonical correlations other than the first. The reason is that a simple permutation scheme does not ignore the variability already explained by previous canonical variables. Here we propose solutions for both problems: in the case of nuisance variables, we show that transforming the residuals to a lower dimensional basis where exchangeability holds results in a valid permutation test; for more general cases, with or without nuisance variables, we propose estimating the canonical correlations in a stepwise manner, removing at each iteration the variance already explained, while dealing with different number of variables in both sides. We also discuss how to address the multiplicity of tests, proposing an admissible test that is not conservative, and provide a complete algorithm for permutation inference for CCA.

  • Attention shifting and subliminal cueing under high attentional load: an EEG study using emotional faces

    by Eda Tipura, Olivier Renaud, Alan J. Pegna

    Emotional expressions and attentional processes interact such that emotion increases perceptual sensitivity in modality-specific brain areas leading to more efficient processing of the stimulus. In the current study, we asked the question of whether the emotional expressions can be processed subliminally when faces are attended but the emotional expression is irrelevant. We presented angry and neutral male and female faces supraliminally and subliminally to participants in an attention-shifting paradigm. Emotional expression was irrelevant to the task, but the gender of the face was predictive of the location of a subsequent target. Analysis of reaction times revealed that, in the supraliminal condition, targets appearing at validly cued locations produced quicker responses than targets at invalidly cued locations. Target-locked event-related potentials (ERPs) confirmed this result and showed that attention selection elicited higher responses at attended locations, as reflected by the N1 component. However, voluntary attention shifting was not triggered when the cue was not consciously perceived. No effects of emotion were observed following target presentation, suggesting that the high attentional load required by the experiment may have prevented the processing of emotion for undetectable faces, as demonstrated by the fact that, following cue onset, the N170 was modulated by emotion in supraliminal trials only. We therefore show that emotional expressions cannot be processed without awareness when faces are attended, but the emotional expression is irrelevant.


  • Interaction and threshold effects of appraisal on componential patterns of emotion: A study using cross-cultural semantic data.

    by Ben Meuleman, Agnes Moors, Johnny Fontaine, Olivier Renaud, Klaus Scherer

    Studies that investigated the relation between appraisal and emotion have largely focused on the linear effect of appraisal criteria on subjective feelings (e.g., the effect of appraised goal obstruction on anger). Emotional responding can be extended to include more than just feelings, however. Componential definitions of emotion also add motivation, physiology, and expression. Moreover, a linear model is not compatible with the idea held by many appraisal theorists that appraisal criteria interact to produce emotional responding. In the present study, we modeled adaptive nonlinear interaction effects of appraisal criteria on motivation, expression, and physiology simultaneously. We applied a combination of principal component analysis for data reduction and multivariate adaptive regression splines (MARS) for automatic interaction identification. Data were obtained from a large-scale cross-cultural study on emotion concepts conducted in 27 countries, which represented semantic profiles of component information in 24 common emotion words. Results of modeling indicated that (a) appraisal of relevance, familiarity, goal compatibility, coping potential, and suddenness showed main effects on component responses; (b) appraisals of agency and norm compatibility uniquely showed interaction effects on component responses; (c) interaction effects explained significant variance only in some component responses but not all; and (d) the emotion patterns simulated by the fitted MARS model could be clustered according to qualitative emotion categories. (PsycINFO Database Record (c) 2019 APA, all rights reserved).