Removal of artifacts from resting-state fMRI data in stroke
We examined the effect of lesion on the resting-state functional connectivity in chronic post-stroke patients. We found many instances of strong correlations in BOLD signal measured at different locations within the lesion, making it hard to distinguish from the connectivity between intact and stron...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2018-01-01
|
| Series: | NeuroImage: Clinical |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S221315821730267X |
| _version_ | 1811274486453895168 |
|---|---|
| author | Grigori Yourganov Julius Fridriksson Brielle Stark Christopher Rorden |
| author_facet | Grigori Yourganov Julius Fridriksson Brielle Stark Christopher Rorden |
| author_sort | Grigori Yourganov |
| collection | DOAJ |
| description | We examined the effect of lesion on the resting-state functional connectivity in chronic post-stroke patients. We found many instances of strong correlations in BOLD signal measured at different locations within the lesion, making it hard to distinguish from the connectivity between intact and strongly connected regions. Regression of the mean cerebro-spinal fluid signal did not alleviate this problem. The connectomes computed by exclusion of lesioned voxels were not good predictors of the behavioral measures. We came up with a novel method that utilizes Independent Component Analysis (as implemented in FSL MELODIC) to identify the sources of variance in the resting-state fMRI data that are driven by the lesion, and to remove this variance. The resulting functional connectomes show better correlations with the behavioral measures of speech and language, and improve the out-of-sample prediction accuracy of multivariate analysis. We therefore advocate this preprocessing method for studies of post-stroke functional connectivity, particularly in samples with large lesions. Keywords: Stroke, fMRI, Functional connectivity, Preprocessing, Independent component analysis, Multivariate prediction |
| first_indexed | 2024-04-12T23:21:03Z |
| format | Article |
| id | doaj.art-ee3d77e5ec5e40fd9d62b727b12ead42 |
| institution | Directory Open Access Journal |
| issn | 2213-1582 |
| language | English |
| last_indexed | 2024-04-12T23:21:03Z |
| publishDate | 2018-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | NeuroImage: Clinical |
| spelling | doaj.art-ee3d77e5ec5e40fd9d62b727b12ead422022-12-22T03:12:32ZengElsevierNeuroImage: Clinical2213-15822018-01-0117297305Removal of artifacts from resting-state fMRI data in strokeGrigori Yourganov0Julius Fridriksson1Brielle Stark2Christopher Rorden3Department of Psychology, University of South Carolina, Columbia, SC 29208, United States; Corresponding author.Department of Communication Science & Disorders, University of South Carolina, Columbia, SC 29208, United StatesDepartment of Communication Science & Disorders, University of South Carolina, Columbia, SC 29208, United StatesDepartment of Psychology, University of South Carolina, Columbia, SC 29208, United StatesWe examined the effect of lesion on the resting-state functional connectivity in chronic post-stroke patients. We found many instances of strong correlations in BOLD signal measured at different locations within the lesion, making it hard to distinguish from the connectivity between intact and strongly connected regions. Regression of the mean cerebro-spinal fluid signal did not alleviate this problem. The connectomes computed by exclusion of lesioned voxels were not good predictors of the behavioral measures. We came up with a novel method that utilizes Independent Component Analysis (as implemented in FSL MELODIC) to identify the sources of variance in the resting-state fMRI data that are driven by the lesion, and to remove this variance. The resulting functional connectomes show better correlations with the behavioral measures of speech and language, and improve the out-of-sample prediction accuracy of multivariate analysis. We therefore advocate this preprocessing method for studies of post-stroke functional connectivity, particularly in samples with large lesions. Keywords: Stroke, fMRI, Functional connectivity, Preprocessing, Independent component analysis, Multivariate predictionhttp://www.sciencedirect.com/science/article/pii/S221315821730267X |
| spellingShingle | Grigori Yourganov Julius Fridriksson Brielle Stark Christopher Rorden Removal of artifacts from resting-state fMRI data in stroke NeuroImage: Clinical |
| title | Removal of artifacts from resting-state fMRI data in stroke |
| title_full | Removal of artifacts from resting-state fMRI data in stroke |
| title_fullStr | Removal of artifacts from resting-state fMRI data in stroke |
| title_full_unstemmed | Removal of artifacts from resting-state fMRI data in stroke |
| title_short | Removal of artifacts from resting-state fMRI data in stroke |
| title_sort | removal of artifacts from resting state fmri data in stroke |
| url | http://www.sciencedirect.com/science/article/pii/S221315821730267X |
| work_keys_str_mv | AT grigoriyourganov removalofartifactsfromrestingstatefmridatainstroke AT juliusfridriksson removalofartifactsfromrestingstatefmridatainstroke AT briellestark removalofartifactsfromrestingstatefmridatainstroke AT christopherrorden removalofartifactsfromrestingstatefmridatainstroke |