ARIbrain-SPM: True discovery proportion-based cluster inference using All-Resolutions Inference
Poster No:
1049
Submission Type:
Abstract Submission
Authors:
Xu Chen1, Wouter Weeda2, Jelle Goeman3
Institutions:
1University of Essex, Colchester, Essex, 2Leiden University, Leiden, South Holland, 3Leiden University Medical Center, Leiden, South Holland
First Author:
Co-Author(s):
Introduction:
The issue of low spatial specificity associated with the popular cluster inference in neuroimaging has been well recognised (Woo et al., 2014; Eklund et al., 2016). That is, for each detected cluster, there exists at least one truly activated voxel, but the exact location and amount of activation within the cluster are unknown. To quantify the amount of signal, several approaches have been proposed by estimating the true discovery proportion (TDP) (Rosenblatt et al., 2018; Blain et al., 2022; Andreella et al., 2023; Vesely et al., 2023; Goeman et al., 2023). Our initial implementation required researchers to provide a cluster for which the TDP was then estimated. We recently developed an adaptive thresholding algorithm which returns clusters with a priori specified TDP (Chen et al. 2023). This greatly improves the usability of the All-Resolutions Inference (ARI) framework, as it simplifies analysis and interpretability.
The free and open-source Statistical Parametric Mapping (SPM) software (Penny et al., 2007), which was designed for the analyses of brain imaging data, is one of the most widely used software packages in this area. The toolboxes that are developed based on SPM are recognised as SPM extensions, which enable the SPM users to carry out specific inferences customised by the developers.
Here we introduce a new SPM extension "ARIbrain-SPM" that can be used to implement the TDP-based cluster inference using ARI, aiming to deal with the problem of low spatial specificity in cluster inference.
The free and open-source Statistical Parametric Mapping (SPM) software (Penny et al., 2007), which was designed for the analyses of brain imaging data, is one of the most widely used software packages in this area. The toolboxes that are developed based on SPM are recognised as SPM extensions, which enable the SPM users to carry out specific inferences customised by the developers.
Here we introduce a new SPM extension "ARIbrain-SPM" that can be used to implement the TDP-based cluster inference using ARI, aiming to deal with the problem of low spatial specificity in cluster inference.
Methods:
Using ARI, the TDP lower bound, which assesses the proportion of active voxels within a cluster, can be measured for any arbitrary cluster/region. If there are no target regions, the adaptive thresholding algorithm can be applied to find all maximal supra-threshold clusters (STCs), each of which has a TDP equal to or exceeding the given threshold.
Therefore, the ARIbrain-SPM toolbox offers two types of TDP-based cluster inference:
(1) clusterTDP (activation quantification) - For significant clusters found using standard cluster inference embedded in SPM, the corresponding TDP lower bounds are estimated using ARI.
(2) tdpCluster (activation localisation) - Given a sufficient TDP threshold, maximal STCs are identified using the adaptive thresholding algorithm. Here, each STC has the TDP greater than or equal to the threshold.
As demonstration, we randomly selected 20 subjects from an fMRI dataset consisting of 218 healthy subjects, where the participants were asked to distinguish between vocal and non-vocal sounds during the experiment (Pernet et al., 2015). The Vocal>Non-vocal contrasts were derived using SPM12 (the latest version of SPM; see more details at https://www.fil.ion.ucl.ac.uk/spm/software/spm12/) for group inference.
Therefore, the ARIbrain-SPM toolbox offers two types of TDP-based cluster inference:
(1) clusterTDP (activation quantification) - For significant clusters found using standard cluster inference embedded in SPM, the corresponding TDP lower bounds are estimated using ARI.
(2) tdpCluster (activation localisation) - Given a sufficient TDP threshold, maximal STCs are identified using the adaptive thresholding algorithm. Here, each STC has the TDP greater than or equal to the threshold.
As demonstration, we randomly selected 20 subjects from an fMRI dataset consisting of 218 healthy subjects, where the participants were asked to distinguish between vocal and non-vocal sounds during the experiment (Pernet et al., 2015). The Vocal>Non-vocal contrasts were derived using SPM12 (the latest version of SPM; see more details at https://www.fil.ion.ucl.ac.uk/spm/software/spm12/) for group inference.
Results:
The results of ARIbrain-SPM can be visualised with the SPM graphical user interface, where the summary table is highly related to the result table of SPM12. Here, a newly added variable "TDP" at the cluster level shows the estimated TDP lower bounds. Figures 1 & 2 illustrate such summary tables on analysing the subsample of the auditory dataset by making clusterTDP and tdpCluster inferences, respectively.
Conclusions:
We have proposed a new SPM extension "ARIbrain-SPM" for conducting the TDP-based cluster inference using ARI in neuroimaging, which is freely accessible at https://github.com/xuchen312/ARIbrain-SPM. With ARIbrain-SPM, users can either measure the amount of activation using clusterTDP, or detect the location of activation using tdpCluster, which largely eases the problem of low spatial specificity in cluster inference.
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI) 1
Methods Development
Univariate Modeling
Other Methods 2
Keywords:
FUNCTIONAL MRI
Open-Source Code
Statistical Methods
Other - All-Resolutions Inference (ARI), true discovery proportion (TDP), cluster inference, SPM extension, spatial specificity, localise activation, quantify activation
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
[2] Blain, A., Thirion, B. and Neuvial, P. (2022). Notip: Non-parametric true discovery proportion control for brain imaging. NeuroImage, 260, 119492.
[3] Chen, X., Goeman, J.J., Krebs, T.J.P., Meijer, R.J. and Weeda, W.D. (2023). Adaptive Cluster Thresholding with Spatial Activation Guarantees Using All-resolutions Inference. arXiv, 2206.13587.
[4] Eklund, A., Nichols, T.E. and Knutsson, H. (2016). Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates. Proceedings of the National Academy of Sciences, 113(28), 7900-7905.
[5] Goeman, J.J., Górecki, P., Monajemi, R., Chen, X., Nichols, T.E. and Weeda, W. (2023). Cluster extent inference revisited: quantification and localisation of brain activity. Journal of the Royal Statistical Society Series B: Statistical Methodology, 85(4), 1128-1153.
[6] Penny, W.D., Friston, K.J., Ashburner, J.T., Kiebel, S.J. and Nichols, T.E. (2007). Statistical Parametric Mapping: The Analysis of Functional Brain Images. Elsevier.
[7] Pernet, C.R., McAleer, P., Latinus, M., Gorgolewski, K.J., Charest, I., Bestelmeyer, P.E.G., Watson, R.H., Fleming, D., Crabbe, F., Valdes-Sosa, M. and Belin, P. (2015). The human voice areas: spatial organization and inter-individual variability in temporal and extra-temporal cortices. NeuroImage, 119, 164-174.
[8] Rosenblatt, J.D., Finos, L., Weeda, W.D., Solari, A. and Goeman, J.J. (2018). All-Resolutions Inference for brain imaging. NeuroImage, 181, 786-796.
[9] Vesely, A., Finos, L. and Goeman, J.J. (2023), Permutation-based true discovery guarantee by sum tests. Journal of the Royal Statistical Society Series B: Statistical Methodology, 85(3), 664-683.
[10] Woo, C.W., Krishnan, A. and Wager, T.D. (2014). Cluster-extent based thresholding in fMRI analyses: pitfalls and recommendations. NeuroImage, 91, 412-419.