A practical guide for improving transparency and reproducibility in neuroimaging research

Gorgolewski, Krzysztof J.; Poldrack, Russell A.

doi:10.1101/039354

Cited by 21 publications

(19 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We have also shown that our results are robust to revised recommendations for minimum sample sizes in CBMAs. Although CBMAs are currently the primary tool for assessing the consistency and specificity of neuroimaging results, image-based metaanalyses [Salimi-Khorshidi et al, 2009] may eventually rise in popularity as more authors share unthresholded statistical maps and adopt more open science practices [Gorgolewski and Poldrack, 2016;McKiernan et al, 2016]. Within the context of PPI studies, image-based meta-analyses would have the power to detect subthreshold connectivity patterns across studies, thus improving our understanding of how brain connectivity shapes behavior.…”

Section: Dorsolateral Prefrontal Cortex (Dlpfc) Target In Fig-mentioning

confidence: 99%

Meta‐analysis of psychophysiological interactions: Revisiting cluster‐level thresholding and sample sizes

Smith

Delgado

2016

Human Brain Mapping

View full text Add to dashboard Cite

Within the neuroimaging community, coordinate based meta-analyses (CBMAs) are essential for aggregating findings across studies and testing whether those studies report similar anatomical locations. This approach has been predominantly applied to studies that focus on whether activation of a brain region is associated with a given psychological process. In a recent paper, we used CBMA to examine a distinct set of studies—i.e., those focusing on whether connectivity between brain regions is modulated by a given psychological process (Smith et al., 2016). Specifically, we reviewed 284 studies examining brain connectivity psychophysiological interactions (PPI). Our meta-analytic results indicated that PPI yields connectivity patterns that are consistent across studies and that can be specific for a given psychological process and seed region. After publication of our findings, we learned that the analysis software we used to conduct our CBMAs (GingerALE v2.3.3) contained an implementation error that led to results that were more liberal than intended. Here, we comment on the impact of this implementation error on the results of our paper, new recommendations for sample sizes in CBMAs, and the importance of communication between software users and developers. We show that our key claims are supported in a reanalysis and that our results are robust to new guidelines on sample sizes.

show abstract

Section: Dorsolateral Prefrontal Cortex (Dlpfc) Target In Fig-mentioning

confidence: 99%

Meta‐analysis of psychophysiological interactions: Revisiting cluster‐level thresholding and sample sizes

Smith

Delgado

2016

Human Brain Mapping

View full text Add to dashboard Cite

show abstract

“…This is exacerbated by the cost of data acquisition and the wide variety and complexity of analyses in the neuroimaging field. In response, practical guidelines for replication in neuroimaging studies have been published (Bakken, 2019;Gorgolewski & Poldrack, 2016), journals have accommodated replication studies (Picciotto, 2018), for example, the Human Brain Mapping Replication Award and the creation of a replication category in NeuroImage: Clinical (Fletcher & Grafton, 2013), and an educational course to teach computation reproducibility has been trialled (Millman, Brett, Barnowski, & Poline, 2018). Successful application of replication analysis principles has provided key advances in the neuroimaging of speech perception (Evans, 2017).…”

Section: Discussionmentioning

confidence: 99%

Replicable brain signatures of emotional bias and memory based on diffusion kurtosis imaging of white matter tracts

Welton

Indja

Maller

et al. 2019

Human Brain Mapping

View full text Add to dashboard Cite

Diffusion MRI (dMRI) is sensitive to anisotropic diffusion within bundles of nerve axons and can be used to make objective measurements of brain networks. Many brain disorders are now recognised as being caused by network dysfunction or are secondarily associated with changes in networks. There is therefore great potential in using dMRI measures that reflect network integrity as a future clinical tool to help manage these conditions. Here, we used dMRI to identify replicable, robust and objective markers that meaningfully reflect cognitive and emotional performance. Using diffusion kurtosis analysis and a battery of cognitive and emotional tests, we demonstrated strong relationships between white matter structure across networks of anatomically and functionally specific brain regions with both emotional bias and emotional memory performance in a large healthy cohort. When the connectivity of these regions was examined using diffusion tractography, the terminations of the identified tracts overlapped precisely with cortical loci relating to these domains, drawn from an independent spatial meta‐analysis of available functional neuroimaging literature. The association with emotional bias was then replicated using an independently acquired healthy cohort drawn from the Human Connectome Project. These results demonstrate that, even in healthy individuals, white matter dMRI structural features underpin important cognitive and emotional functions. Our robust cross‐correlation and replication supports the potential of structural brain biomarkers from diffusion kurtosis MRI to characterise early neurological changes and risk in individuals with a reduced threshold for cognitive dysfunction, with further testing required to demonstrate clinical utility.

show abstract

“…In fact, recent years have seen an increase in alarming signals regarding the lack of replicability in neuroimaging research [13,14]. There are many factors influencing this problem, some subjectrelated such as the variability in the collection of phenotypic data and the inherent heterogeneity in disease, but also the acquisition hardware and data analysis methods can play a major role.…”

Section: Discussionmentioning

confidence: 99%

Quantitative Intensity Harmonization of Dopamine Transporter SPECT Images Using Gamma Mixture Models

et al. 2018

View full text Add to dashboard Cite

Purpose: Differences in site, device, and/or settings may cause large variations in the intensity profile of dopamine transporter (DAT) single-photon emission computed tomography (SPECT) images. However, the current standard to evaluate these images, the striatal binding ratio (SBR), does not efficiently account for this heterogeneity and the assessment can be unequivalent across distinct acquisition pipelines. In this work, we present a voxel-based automated approach to intensity normalize such type of data that improves on cross-session interpretation. Procedures: The normalization method consists of a reparametrization of the voxel values based on the cumulative density function (CDF) of a Gamma distribution modeling the specific region intensity. The harmonization ability was tested in 1342 SPECT images from the PPMI repository, acquired with 7 distinct gamma camera models and at 24 different sites. We compared the striatal quantification across distinct cameras for raw intensities, SBR values, and after applying the Gamma CDF (GDCF) harmonization. As a proof-of-concept, we evaluated the impact of GCDF normalization in a classification task between controls and Parkinson disease patients. Results: Raw striatal intensities and SBR values presented significant differences across distinct camera models. We demonstrate that GCDF normalization efficiently alleviated these differences in striatal quantification and with values constrained to a fixed interval [0,1]. Also, our method allowed a fully automated image assessment that provided maximal classification ability, given by an area under the curve (AUC) of AUC = 0.94 when used mean regional variables and AUC = 0.98 when used voxel-based variables. Conclusion: The GCDF normalization method is useful to standardize the intensity of DAT SPECT images in an automated fashion and enables the development of unbiased algorithms using multicenter datasets. This method may constitute a key pre-processing step in the analysis of this type of images.Alberto Llera and Ismael Huertas shared first authorship. Electronic supplementary material The online version of this article (https://

show abstract

A practical guide for improving transparency and reproducibility in neuroimaging research

Cited by 21 publications

References 37 publications

Meta‐analysis of psychophysiological interactions: Revisiting cluster‐level thresholding and sample sizes

Meta‐analysis of psychophysiological interactions: Revisiting cluster‐level thresholding and sample sizes

Replicable brain signatures of emotional bias and memory based on diffusion kurtosis imaging of white matter tracts

Quantitative Intensity Harmonization of Dopamine Transporter SPECT Images Using Gamma Mixture Models

Contact Info

Product

Resources

About