Assessing the Reliability of fMRI-CPCA for Single-Subject Functional Brain Network Detection
Poster No:
1901
Submission Type:
Abstract Submission
Authors:
Sofia Eickhoff1, Erica Zeng2, Abhijit Chinchani2, Ava Momeni2, Christina Andreou1, Stefan Borgwardt1, Todd Woodward2
Institutions:
1University of Luebeck, Luebeck, Germany, 2University of British Columbia, Vancouver, Canada
First Author:
Co-Author(s):
Introduction:
Dysfunctions in task-based brain networks measured through functional magnetic resonance imaging (fMRI) are increasingly associated with cognitive and behavioral deficits in schizophrenia (SZ) (Barch, 2007; Fouladirad, 2022; Goghari, 2017; Lavigne, 2020). While group-level analyses of task-based fMRI data have provided valuable insights, clinical translation requires validity at the individual level. Our in-house software, constrained principal component analysis for fMRI (fMRI-CPCA), offers a promising method for extracting blood-oxygen level-dependent (BOLD) signals without relying on assumed task models, which can introduce noise when mismatched with actual BOLD changes (Metzak, 2011; Momeni, 2024, Sanford, 2020; Woodward, 2006). By isolating signals through dimensional reduction, this technique holds potential for personalized diagnostics and treatments in schizophrenia.
Our objective was to take a first step in testing the reliability of fMRI-CPCA for detecting task-related functional brain networks at the single-subject level. Specifically, we aimed to assess how well individual networks align with group-level patterns and highlight the need for further investigations into detecting abnormalities in SZ patients.
Our objective was to take a first step in testing the reliability of fMRI-CPCA for detecting task-related functional brain networks at the single-subject level. Specifically, we aimed to assess how well individual networks align with group-level patterns and highlight the need for further investigations into detecting abnormalities in SZ patients.
Methods:
Task-based fMRI data from a probabilistic reasoning task (N=41 controls, N=71 patients) that was previously published by our laboratory were analyzed (Fouladirad, 2022). Single-subject analyses were performed using fMRI-CPCA, applying the group mask and extracting the same number of components as the group analysis to facilitate comparisons. Reliability was evaluated by calculating cosine similarity between spatial components derived from single-subject analyses and group-level networks. Singular Value Decomposition (SVD) was used to decompose fMRI data, with right singular vectors representing spatial patterns of interest. A 4x4 cosine similarity matrix was produced for each subject, categorizing matches as good (>0.7), moderate (0.4–0.7), or poor (<0.4).
·fMRI-CPCA Analysis Pipeline
Results:
Cosine similarity analysis showed that 38% of individual components had weak matches (<0.4), 40% demonstrated moderate matches (0.4–0.7), and 22% exhibited strong matches (>0.7) with group-level components. Group component 1 (sustained visual attention network) consistently achieved the strongest alignment, with 62% of subjects displaying moderate-to-strong matches, suggesting its robustness. In contrast, components 2 (response network) and 3 (auditory-attention-for-response network) displayed greater variability, with a higher proportion of weak matches. Group component C4 (sustained visual attention/default mode network) demonstrated intermediate stability, with 45% of subjects achieving moderate-to-strong matches.
·Cosine Similarity Matches with Group-Level Components
Conclusions:
These findings provide a first, promising impression of the reliability of fMRI-CPCA in detecting task-related functional brain networks at the single-subject level. Components 1 and 4, which showed the strongest matches, represent similar overall anatomical patterns in group analysis, suggesting that some networks may be more reliably detectable at the individual level than others. However, this study represents only an initial step. Future research must further evaluate the reliability of single-subject results, particularly in terms of their alignment with group-level BOLD changes induced by tasks. Additionally, it remains to be tested whether single-subject analyses can detect abnormalities in SZ patients, a critical step toward clinical application. By capturing individualized network patterns, this method could pave the way for identifying biomarkers and informing precision psychiatry approaches for schizophrenia, including early detection and personalized treatment strategies.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia)
Higher Cognitive Functions:
Higher Cognitive Functions Other
Modeling and Analysis Methods:
Multivariate Approaches 2
Novel Imaging Acquisition Methods:
BOLD fMRI 1
Keywords:
FUNCTIONAL MRI
Multivariate
Psychiatric Disorders
Schizophrenia
Statistical Methods
Other - Task-based Functional Brain Networks, Single-Subject fMRI-CPCA, Reliability Testing
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.
Fouladirad, S. (2022). Functional brain networks underlying probabilistic reasoning and delusions in schizophrenia. Psychiatry Research: Neuroimaging, 323, 111472.
Goghari, V. M. (2017). Task-related functional connectivity analysis of emotion discrimination in a family study of schizophrenia. Schizophrenia Bulletin, 43(6), 1348.
Lavigne, K. M. (2020). Functional brain networks underlying evidence integration and delusional ideation. Schizophrenia Research, 216, 302–309.
Momeni, A. (2024). Functional brain networks underlying autobiographical event simulation: An update. PsyArXiv.
Metzak, P. (2011). Constrained principal component analysis reveals functionally connected load‐dependent networks involved in multiple stages of working memory. Human Brain Mapping, 32(6), 856-871.
Sanford, N. (2020). Task-merging for finer separation of functional brain networks in working memory. Cortex, 125, 246-271.
Woodward, T. S. (2006). Functional connectivity reveals load dependent neural systems underlying encoding and maintenance in verbal working memory. Neuroscience, 139(1), 317-325.