Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear Model
Group-level repeated measurements are common in neuroimaging, yet their analysis remains complex. Although a variety of specialized tools now exist, it is surprising that to-date there has been no clear discussion of how repeated-measurements can be analyzed appropriately using the standard general...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2019-04-01
|
| Series: | Frontiers in Neuroscience |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/article/10.3389/fnins.2019.00352/full |
| _version_ | 1828802497401061376 |
|---|---|
| author | Martyn McFarquhar |
| author_facet | Martyn McFarquhar |
| author_sort | Martyn McFarquhar |
| collection | DOAJ |
| description | Group-level repeated measurements are common in neuroimaging, yet their analysis remains complex. Although a variety of specialized tools now exist, it is surprising that to-date there has been no clear discussion of how repeated-measurements can be analyzed appropriately using the standard general linear model approach, as implemented in software such as SPM and FSL. This is particularly surprising given that these implementations necessitate the use of multiple models, even for seemingly conventional analyses, and that without care it is very easy to specify contrasts that do not correctly test the effects of interest. Despite this, interest in fitting these types of models using conventional tools has been growing in the neuroimaging community. As such it has become even more important to elucidate the correct means of doing so. To begin, this paper will discuss the key concept of the expected mean squares (EMS) for defining suitable F-ratios for testing hypotheses. Once this is understood, the logic of specifying correct repeated measurements models in the GLM should be clear. The ancillary issue of specifying suitable contrast weights in these designs will also be discussed, providing a complimentary perspective on the EMS. A set of steps will then be given alongside an example of specifying a 3-way repeated-measures ANOVA in SPM. Equivalency of the results compared to other statistical software will be demonstrated. Additional issues, such as the inclusion of continuous covariates and the assumption of sphericity, will also be discussed. The hope is that this paper will provide some clarity on this confusing topic, giving researchers the confidence to correctly specify these forms of models within traditional neuroimaging analysis tools. |
| first_indexed | 2024-12-12T07:05:15Z |
| format | Article |
| id | doaj.art-2f62a14e593d406c860b821d866729de |
| institution | Directory Open Access Journal |
| issn | 1662-453X |
| language | English |
| last_indexed | 2024-12-12T07:05:15Z |
| publishDate | 2019-04-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neuroscience |
| spelling | doaj.art-2f62a14e593d406c860b821d866729de2022-12-22T00:33:45ZengFrontiers Media S.A.Frontiers in Neuroscience1662-453X2019-04-011310.3389/fnins.2019.00352414497Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear ModelMartyn McFarquharGroup-level repeated measurements are common in neuroimaging, yet their analysis remains complex. Although a variety of specialized tools now exist, it is surprising that to-date there has been no clear discussion of how repeated-measurements can be analyzed appropriately using the standard general linear model approach, as implemented in software such as SPM and FSL. This is particularly surprising given that these implementations necessitate the use of multiple models, even for seemingly conventional analyses, and that without care it is very easy to specify contrasts that do not correctly test the effects of interest. Despite this, interest in fitting these types of models using conventional tools has been growing in the neuroimaging community. As such it has become even more important to elucidate the correct means of doing so. To begin, this paper will discuss the key concept of the expected mean squares (EMS) for defining suitable F-ratios for testing hypotheses. Once this is understood, the logic of specifying correct repeated measurements models in the GLM should be clear. The ancillary issue of specifying suitable contrast weights in these designs will also be discussed, providing a complimentary perspective on the EMS. A set of steps will then be given alongside an example of specifying a 3-way repeated-measures ANOVA in SPM. Equivalency of the results compared to other statistical software will be demonstrated. Additional issues, such as the inclusion of continuous covariates and the assumption of sphericity, will also be discussed. The hope is that this paper will provide some clarity on this confusing topic, giving researchers the confidence to correctly specify these forms of models within traditional neuroimaging analysis tools.https://www.frontiersin.org/article/10.3389/fnins.2019.00352/fullrepeated measurementswithin-subjectflexible factorialSPMFSLGLM |
| spellingShingle | Martyn McFarquhar Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear Model Frontiers in Neuroscience repeated measurements within-subject flexible factorial SPM FSL GLM |
| title | Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear Model |
| title_full | Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear Model |
| title_fullStr | Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear Model |
| title_full_unstemmed | Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear Model |
| title_short | Modeling Group-Level Repeated Measurements of Neuroimaging Data Using the Univariate General Linear Model |
| title_sort | modeling group level repeated measurements of neuroimaging data using the univariate general linear model |
| topic | repeated measurements within-subject flexible factorial SPM FSL GLM |
| url | https://www.frontiersin.org/article/10.3389/fnins.2019.00352/full |
| work_keys_str_mv | AT martynmcfarquhar modelinggrouplevelrepeatedmeasurementsofneuroimagingdatausingtheunivariategenerallinearmodel |