Manuscript

2025

1. Cerebral Cortex

   

   Structural mediation of the default-mode network in children with callosal agenesis

   Céline Provins, Anjali Tarun Nahalka, Léa Schmidt, and 8 more authors

   Cerebral Cortex , Jul 2025

   Abs HTML

   

   Agenesis of the corpus callosum is a neurodevelopmental condition characterized by the partial or complete absence of the corpus callosum, the largest white matter bundle connecting the cerebral hemispheres. The default-mode network comprises bilateral frontal, temporal, and parietal regions that exhibit correlated activity at rest. Previous studies show that individuals with agenesis of the corpus callosum show overall preserved default-mode network functional connectivity, suggesting compensatory mechanisms for maintaining bilaterally correlated activity. In this study, we aimed to explore white matter pathways that support default-mode network-related networks in 15 children with agenesis of the corpus callosum and 27 typically developing controls, using combined diffusion and functional magnetic resonance imaging. A seed-based and dynamic functional connectivity approach enabled us to examine default-mode network spatial and temporal properties and their white matter substrates. While spatial default-mode network patterns were similar across groups, we found differences in temporal dynamics of 1 network and in white matter–default-mode network correspondence. These differences were either observed in white matter tracts directly associated with complete or partial absence of the corpus callosum or in white matter tracts such as the fornix and the anterior and posterior commissures, which have been previously implicated in neuroplasticity in agenesis of the corpus callosum. Our findings show that default-mode network dynamics can remain functionally preserved despite significant white matter alterations.

2023

1. Front. Neuroimaging

   

   Quality control in functional MRI studies with MRIQC and fMRIPrep

   Céline Provins, Eilidh MacNicol, Saren H. Seeley, and 2 more authors

   Frontiers in Neuroimaging , Jul 2023

   Abs HTML

   

   The implementation of adequate quality assessment (QA) and quality control (QC) protocols within the magnetic resonance imaging (MRI) research workflow is resource- and time-consuming and even more so is their execution. As a result, QA/QC practices highly vary across laboratories and “MRI schools”, ranging from highly specialized knowledge spots to environments where QA/QC is considered overly onerous and costly despite evidence showing that below-standard data increase the false positive and false negative rates of the final results. Here, we demonstrate a protocol based on the visual assessment of images one-by-one with reports generated by MRIQC and fMRIPrep, for the QC of data in functional (blood-oxygen dependent-level; BOLD) MRI analyses. We particularize the proposed, open-ended scope of application to whole-brain voxel-wise analyses of BOLD to correspondingly enumerate and define the exclusion criteria applied at the QC checkpoints. We apply our protocol on a composite dataset (n = 181 subjects) drawn from open fMRI studies, resulting in the exclusion of 97% of the data (176 subjects). This high exclusion rate was expected because subjects were selected to showcase artifacts. We describe the artifacts and defects more commonly found in the dataset that justified exclusion. We moreover release all the materials we generated in this assessment and document all the QC decisions with the expectation of contributing to the standardization of these procedures and engaging in the discussion of QA/QC by the community.

2. Nature Methods

   

   Reliability characterization of MRI measurements for analyses of brain networks on a human phantom

   Céline Provins, Hélène Lajous, Elodie Savary, and 7 more authors

   Stage 1 Registered Report accepted in principle at Nature Methods , Jul 2023

   Abs HTML

   Network-based approaches are widely adopted to model functional and structural ‘connectivity’ of the living brain, extracted noninvasively with magnetic resonance imaging (MRI). However, these analyses —on functional and structural networks— render unreliable at the finer temporal, spatial, and brain-parcellation scales. Consequently, the clinical translation of these analyses has yet to materialize meaningfully, and interpretation of the skyrocketing production of scientific literature requires caution. We will characterize relevant sources of variability and assess the reliability of structural and functional networks extracted from MRI with the repeated acquisition of a single, healthy individual, whom we regard as the ‘Human Connectome Phantom’. Two comprehensive MRI protocols will be executed across three different devices ( 48, 12, and 12 sessions, respectively) while recording a wealth of physiological signals to help model corresponding spurious effects on brain networks. To maximize reuse, e.g., as a benchmark reference, a baseline for machine learning models, or a source of prior knowledge, we will openly share all data and their derivatives. By systematically assessing spurious sources of variability throughout the neuroimaging workflow, we will deliver reliability margins of brain networks that inform future research and contribute to the standardization of ‘connectivity measurement’.

3. PLOS Biology

   

   Defacing biases in manual and automatic quality assessments of structural MRI with MRIQC

   Céline Provins, Elodie Savary, Thomas Sanchez, and 8 more authors

   Stage 1 Registered Report accepted in principle at PLOS Biology , Jul 2023

   Abs HTML

   A critical requirement before data-sharing of human neuroimaging is removing facial features to protect individuals’ privacy. However, not only does this process redact identifiable information about individuals, but it also removes non-identifiable information. This introduces undesired variability into downstream analysis and interpretation. This registered report investigated the degree to which the so-called defacing altered the quality assessment of T1-weighted images of the human brain from the openly available “IXI dataset”. The effect of defacing on manual quality assessment was investigated on a single-site subset of the dataset (N=185). By comparing two linear mixed-effects models, we determined that four trained human raters’ perception of quality was significantly influenced by defacing by comparing their ratings on the same set of images in two conditions: “nondefaced” (that is, preserving facial features) and “defaced”. In addition, we investigated these biases on automated quality assessments by applying repeated-measures multivariate ANOVA (rm-MANOVA) on the image quality metrics extracted with MRIQC on the full IXI dataset (N=581; three acquisition sites). This study found that defacing altered the quality assessments by humans and showed that MRIQC’s quality metrics were mostly insensitive to defacing.

Thesis

2025

1. Doctoral Thesis

   

   Characterizing and Improving the Reliability of Brain Connectivity extracted with Magnetic Resonance Imaging

   Céline Provins

   Mar 2025

   Abs HTML

   Connectivity analysis in neuroimaging has grown immensely over the last two decades as understanding structural and functional networks has provided great insights into brain function, structure, and disease. Although MRI is a versatile imaging modality to explore these networks, the indirect nature of MRI measurements and its sensitivity to numerous noise sources pose substantial challenges to obtaining reliable measurements. Additionally, to extract structural and functional connectivity from the noise, numerous processing steps are required, which has led to the development of a wealth of neuroimaging workflows, each with its own merits but also methodological nuances. This extensive methodological flexibility further undermines the reliability of brain connectivity results. Collectively, these issues restrict the translation of network-based biomarkers into clinical applications. In this thesis, we aimed to address these challenges by proposing a framework for characterizing and improving the reliability of brain connectivity estimates. We approach the broad and multifaceted problem of reliability characterization by decomposing it into three levels: (1) participant-level variability, (2) scanner-level variability, and (3) data processing-level variability. Moreover, we present four approaches to increasing reliability: redundancy, standardization, quality assurance and quality control (QA/QC), and registered reports. The first project presents how we applied the principles of redundancy and standardization to establish a comprehensive and robust QA/QC protocol. In addition to presenting the protocol, our paper highlighted the importance of setting up several checkpoints throughout the workflow and adapting QA/QC procedures to the specifics of the studies. That protocol, combined with our dataset quality annotations, serves as a stepping stone towards developing automatic algorithms to complement visual QA/QC, alleviating researchers’ workload and correcting human-related inconsistencies. We encourage other researchers to use our protocol as a solid foundation to adapt to their own needs. In our second study, we leveraged our standardized QA/QC protocol to investigate the impact of defacing—a necessary step to protect participant privacy when sharing data publicly—on perceived image quality. We found that defacing significantly altered human perception of quality, suggesting that performing QA/QC on defaced data is suboptimal as it risks failing to remove subpar images that could subsequently bias the results. We therefore advise future studies to perform QA/QC before defacing or any other step that could alter the data integrity. Moreover, QA/QC materials extracted from unaltered data should be included when publicly sharing datasets. Machine perception, on the other hand, was found unaffected by defacing because MRIQC did not take into account areas typically altered by defacing. If the field is to follow the recommendations above, future development of QA/QC tools such as MRIQC should correct this weakness. Building on these principles of redundancy, standardization, and the importance of QA/QC, our third study introduced the Human Connectome Phantom (HCPh) dataset, a novel dataset designed to address key challenges in connectivity reliability. The dataset features repeated scans of a single individual across three scanners using multi-echo functional MRI (fMRI) and multi-shell diffusion MRI. This design provides the unique opportunity to characterize the interscanner reliability of both functional and structural connectivity within a unified framework. The sessions on one scanner also incorporated extensive physiological monitoring and acquired the resting-state fMRI using a naturalistic paradigm to enable a detailed examination of intrascanner variability. We encourage the neuroimaging community to leverage this extremely valuable dataset to evaluate the reliability of a vast range of modeling methods whose reliability could not be characterized before. This dataset also offers unique opportunities to develop new methodologies to remove fMRI signal fluctuations from non-neural sources more efficiently. Overall, this thesis paves the way for improving and better characterizing the reliability of functional and structural connectivity, thereby deepening our understanding of their feasibility as biomarkers.

Conference Proceedings

2025

1. OHBM

   

   MRIQC IQM Variability and Artifact Prediction in Multi-Echo fMRI

   Mélanie Garcia, Céline Provins, and Oscar Esteban

   In 31st Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Brisbane, Jun 2025

   Abs HTML

   

   Assessing BOLD fMRI signal quality is essential for reliable statistical analyses. MRIQC (Esteban et al., 2017) is widely used for single-echo fMRI quality control, extracting image quality metrics (IQMs) per scan and generating visual reports to identify outliers at the group level. For multi-echo (ME) fMRI, MRIQC produces IQMs for each echo, but their variation distributions across echoes and their relationship to artifacts in the final scan remain unclear. This study aims to address these gaps, identify aggregation strategies for IQMs, and predict artifact presence using machine learning. ME-fMRI acquires data at multiple echo times, combining them to improve signal quality (Kundu et al., 2017). This technique has gained traction over the past decade for its ability to better disentangle artifact signals from BOLD signal of interest (DuPre et al., 2021).

2. OHBM

   

   Can Replaying a Movie Clip Reduce Variability in Single-Subject Resting-State fMRI?

   Céline Provins, Alexandre Cionca, Élodie Savary, and 5 more authors

   In 31st Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Brisbane , Jun 2025

   Abs

   Resting-state functional MRI (rs-fMRI) is influenced by ongoing cognition, which is neither measured nor controlled (Finn, 2021). This variability complicates signal interpretation and undermines the reliability of edge-level connectivity estimates (Nakuci et al., 2023; Noble et al., 2017). Naturalistic paradigms using real-world stimuli, like movies, help maintain wakefulness and synchronize brain activity across sessions and individuals through time-locked stimuli (Hasson et al., 2004). In this project, we tested stimulus habituation by repeatedly playing a naturalistic, mute movie clip while acquiring fMRI on a single individual. We quantified variability as residual variance after modeling functional connectivity (FC) with principal component analysis (PCA).

2024

1. OHBM

   

   Physiopy: a Python suite for handling physiological data recorded in neuroimaging settings

   Roza Bayrak, Catie Chang, Katherine L. Bottenhorn, and 26 more authors

   In 30th Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Seoul, Jun 2024

   Abs

   Functional Magnetic Resonance Imaging (fMRI), a pivotal tool for neuroscientific research, leverages blood oxygenation levels to infer neural activity. However, its reliance on hemodynamic responses also renders it sensitive to various physiological processes affecting blood oxygenation. This dual nature presents both challenges and opportunities: while these physiological factors can introduce confounds in interpreting neural signals [1], they simultaneously offer valuable insights into essential human functions encompassing cognition, emotion, and health [2-4]. To this end, we underscore the necessity of acquiring concurrent physiological data such as cardiac and respiratory activity, gas exchange metrics (O2/CO2 levels), and skin conductance. Adoption of concurrent physiological signals is growing within the neuroimaging community, reflecting a broader appreciation of physiological dynamics in brain imaging studies. Emphasizing the critical role of physiological monitoring in fMRI data quality, physiopy is a dynamic, collaborative initiative designed to streamline the integration of physiological data with fMRI research. The foundation of physiopy rests on four key pillars: (1) Accessible Software Suite: Offering a range of user-friendly software tools specifically tailored for efficient physiological data processing, (2) Comprehensive Documentation: Ensuring clarity and ease of use through detailed guides and instructional materials, (3) Community-Driven Practices: Fostering a culture of shared knowledge and collaborative development of best practices, and (4) Engaged Community: Cultivating an active network of users, developers, and researchers, all united by a shared interest in the integration of physiology within neuroimaging research.

2. OHBM

   

   A high definition anatomical brain template of one individual healthy subject

   Alexandre Cionca, Yasser Alemán- Gómez, Céline Provins, and 3 more authors

   In 30th Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Seoul , Jun 2024

   Abs HTML

   

   Brain templates aggregate averaged features across subjects in a normalized stereotactic space and are critical to formalize prior knowledge in neuroimaging analyses. The most widespread templates have been developed by the Montreal Neurological Institute (MNI), yielding a number of templates referred to with the umbrella term of “MNI space”. Recently, some attention shifted towards emphasizing depth rather than width in datasets with few subjects that have many images. Such “dense sampling” protocols, coupled with modern neuroimaging tools, are effective at enhancing image resolution and provide more accurate surfaces. With this in mind, we propose a high-definition template of a single healthy brain built from 70 MR images with multimodal registration

3. ISMRM

   

   Harnessing QA/QC protocols for diffusion MRI neuroimaging workflows with MRIQC

   Yibei Chen, Céline Provins, Christopher J. Markiewicz, and 2 more authors

   In Annual Meeting of the International Society for Magnetic Resonance in Medicine (ISMRM) in Singapore , May 2024

   Abs HTML

   

   Motivation. Reliable neuroimaging pipelines require the implementation of robust QA/QC protocols. Goals. Developing an MRIQC extension for QA/QC of diffusion MRI data. Approach. We build on MRIQC’s infrastructure to generate individual visual reports of dMRI images and define new image quality metrics (IQMs). Results. We developed a minimal processing pipeline for whole-brain dMRI data of human adults. The processing pipeline generates individual visual reports for the QA of unprocessed inputs. The pipeline also extracts IQMs to train automated decision-making, following MRIQC’s established pattern. Impact. MRIQC is a widely-adopted tool for the QA/QC of unprocessed MRI data. However, support for dMRI was previously lacking. This MRIQC extension will improve QA/QC of dMRI by bringing it to the highest standards and will facilitate the implementation of rigorous protocols in multimodal neuroimaging.

4. ISMRM

   

   QA/QC of Anatomical, Functional, and Diffusion MRI Data of the Human Connectome Phantom (HCPh) Dataset with MRIQC and fMRIPrep

   Céline Provins, and Oscar Esteban

   In Annual Meeting of the International Society for Magnetic Resonance in Medicine (ISMRM) in Singapore , May 2024

   HTML

   
5. OHBM

   

   The Human Connectome Phantom (HCPh) dataset

   Oscar Esteban, Céline Provins, and Alexandre Cionca

   In 30th Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Seoul , Jun 2024

   Abs HTML

   Network-based approaches are widely adopted to model functional and structural ‘connectivity’ of the living brain, extracted noninvasively with magnetic resonance imaging (MRI). However, these analyses —on functional and structural networks— render unreliable at the finer temporal, spatial, and brain-parcellation scales. Here, we introduce the Human Connectome PHantom (HCPh) dataset as part of our recent Stage 1 Registered Report1 (Fig. 1).

6. OHBM

   

   Effects of phase-encoding on BOLD data with a positive control task

   Céline Provins, Alexandre Cionca, Elodie Savary, and 2 more authors

   In 30th Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Seoul, Jun 2024

   Abs HTML

   

   Quality assessment and quality control (QA/QC) checkpoints layered throughout the dataflow are essential to ensure the reliability of neuroimaging analyses. In the case of functional MRI, best practices recommend collecting a ‘positive control’ task with which the different layers of QA/QC can be validated. These are short and simple tasks designed to elicit robust and precisely located brain activation patterns, permitting the diagnosis of potential issues in the workflow. Here, we examine how the phase-encoding direction (PE) choice in echo-planar imaging (EPI) blood-oxygen-level-dependent (BOLD) fMRI influences the resulting activation maps using a positive control task that includes visual and motor paradigms.

2023

1. OHBM

   

   Assessment of B1 field dynamics of rats BOLD fMRI using the wavelet transform

   Céline Provins, Eilidh MacNicol, Elodie Savary, and 2 more authors

   In 29th Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Montreal , Jul 2023

   Abs HTML

   

   Magnetic resonance imaging (MRI) generates a radiofrequency field (B1) to frequency encode the object being imaged. Deviations from the nominal B1 field produce artifactual intensity nonuniformity (INU) across the image, which is problematic, especially for automated analyses that assume a tissue is represented by voxels of similar intensity throughout the image (Belaroussi et al. 2006). These artifacts are particularly exacerbated by receiver coil failures. Such events are difficult to capture as they tend to be short-lived and sporadic. In brain blood-oxygen-level-dependent (BOLD) functional MRI (fMRI), B1 field dynamics is usually visualized with a video of the scan to spot signal intensity changes, but this method is time-consuming and error-prone, as the human observer needs to keep focus during the whole video. Here, we showcase a visualization tool to assess B1 field dynamics and a derived summary metric to efficiently detect low spatial-frequency artifacts, such as transient INU.

2. OHBM

   

   Q’kay: a manager for the quality assessment of large neuroimaging studies

   Elodie Savary, Céline Provins, Thomas Sanchez, and 1 more author

   In 29th Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Montreal , Jul 2023

   Abs HTML

   

   Despite substantial efforts toward improving the tools to carry out the visual assessment of quality, as well as automation, the quality control (QC) of imaging data remains an onerous, yet critical step of analysis workflows, especially within large-scale studies. Indeed, the reliability and reproducibility of results can be improved by implementing QC checkpoints throughout the workflow (Niso et al. 2022, Provins et al. 2023). Here, we introduce Q’kay, a web service to deploy rigorous QC protocols on large datasets leveraging the individual reports generated by tools like MRIQC (Esteban et al., 2017) and fMRIPrep (Esteban et al., 2019).

3. OHBM

   

   Signal-to-noise ratio estimates predict head motion presence in T1-weighted MRI

   Céline Provins, Mikkel Schöttner, Michael Dayan, and 10 more authors

   In 29th Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Montreal , Jul 2023

   Abs HTML

   

   MRIQC (Esteban et al. 2017) is a tool to help researchers perform quality control (QC) on their structural and functional MRI data. Not only does MRIQC generate visual reports for reliable, manual assessment but it also automatically extracts a set of image quality metrics (IQMs). However, these IQMs are hard to interpret, and many related questions remain open, such as which IQMs are more important. In this project, which emerged as a BrainHack Geneva 2022 initiative, we show that head motion during the acquisition of whole-brain T1-weighted (T1w) MRI of healthy volunteers can be predicted based on the IQMs using supervised machine learning. To do so, we employ the open MR-ART (movement-related artifacts; Nárai et al. 2022) dataset, which includes T1w images acquired under three different motion conditions. We show that signal-to-noise ratio (SNR) derived metrics are the most important features to predict motion presence.

2022

1. OHBM

   

   Defacing biases manual and automated quality assessments of structural MRI with MRIQC

   Céline Provins, Yasser Alemán-Gómez, Martine Cleusix, and 4 more authors

   In 28th Annual Meeting of the Organization for Human Brain Mapping (OHBM) in Glasgow , Jun 2022

   type: article

   Abs HTML

   

   Defacing (i.e. removing facial features) from structural imaging has become a necessary step before data sharing to ensure participants’ anonymity (Schwarz et al. 2021; Fig 1A). This process has proven to have some deleterious effects on the downstream research workflow (de Sitter et al. 2020). Here, we present an exploratory analysis prior to testing the hypothesis that both quality ratings by human experts and the image quality metrics (IQMs) that MRIQC (Esteban et al. 2017) extracts are affected by defacing. We found sufficient evidence on a small sample that there might be an effect. Therefore, we will pre-register and carry out a confirmatory analysis on a larger, unseen, sample.

2. EMIM

   Extending MRIQC to rodents: image quality metrics for rat MRI

   Eilidh MacNicol,  McKenzie P. Hagen, Céline Provins, and 3 more authors

   In 17th Annual Meeting of the European Society for Molecular Imaging (EMIM) in Thessaloniki, Greece,, Mar 2022

   Abs

   The appeal of MRI for brain research lies in its capacity to non-invasively measure biological changes but, to reliably represent the ground truth, measurements rely on image quality. Image quality metrics (IQMs) can be generated by MRIQC1, a popular tool that automates quality prediction in brain MR images. Until now, MRIQC’s workflows relied on a single reference space while the software tools and algorithms were developed for use with adult human MRI and make assumptions that do not extend to other species and populations (e.g., babies), so MRIQC’s generalisability was limited.

3. ISMRM

   

   Quality control and nuisance regression of fMRI, looking out where signal should not be found

   Céline Provins, Christopher J. Markiewicz, Rastko Ciric, and 5 more authors

   In Annual Meeting of the International Society for Magnetic Resonance in Medicine (ISMRM) in London, May 2022

   Abs HTML

   

   Quality control of functional MRI data is essential as artifacts can have a critical impact on subsequent analysis. Yet, visual assessment of a dataset is tedious and time-consuming. By extending the carpet plot with the voxels located on a closed band (or “crown”) around the brain, we showed that fMRI data quality can be assessed more effectively. This new feature has been incorporated into MRIQC and fMRIPrep. In addition, a new nuisance regressor has been added to the latter, calculated from timeseries within this new “crown”.