Aphasia Neuroplasticity Review

Griffis et al. (2017a)

Reference

AuthorsGriffis JC, Nenert R, Allendorfer JB, Szaflarski JP
TitleLinking left hemispheric tissue preservation to fMRI language task activation in chronic stroke patients
ReferenceCortex 2017a; 96: 1-18
PMID28961522
DOI10.1016/j.cortex.2017.08.031

Participants

LanguageUS English
Inclusion criteria
Number of individuals with aphasia43
Number of control participants43
Were any of the participants included in any previous studies?Yes (same dataset as Griffis et al. (2017) Hum Brain Mapp)
Is age reported for patients and controls, and matched?Yes (mean 53 ± 15 years, range 23-90 years)
Is sex reported for patients and controls, and matched?Yes (males: 25; females: 18)
Is handedness reported for patients and controls, and matched?Yes (right: 41; left: 2)
Is time post stroke onset reported and appropriate to the study design?Yes (range 1-14 years)
To what extent is the nature of aphasia characterized?Not at all
Language evaluationBNT, semantic fluency, phonemic fluency
Aphasia severityNot stated
Aphasia typeNot stated
First stroke only?Yes
Stroke typeNot stated
To what extent is the lesion distribution characterized?Lesion overlay
Lesion extentMean 105.2 ± 76.3 cc
Lesion locationL
Participants notes

Imaging

ModalityfMRI
Is the study cross-sectional or longitudinal?Cross-sectional
If longitudinal, at what time point(s) were imaging data acquired?
If longitudinal, was there any intervention between the time points?
Is the scanner described?No (Siemens Allegra 3 Tesla or Philips 3 Tesla; model not stated)
Is the timing of stimulus presentation and image acquisition clearly described and appropriate?Yes
Design typeBlock
Total images acquired165
Are the imaging acquisition parameters, including coverage, adequately described and appropriate?Yes (whole brain)
Is preprocessing and intrasubject coregistration adequately described and appropriate?Yes
Is first level model fitting adequately described and appropriate?Yes
Is intersubject normalization adequately described and appropriate?Yes
Imaging notes

Conditions

Are the conditions clearly described?Yes
ConditionResponse typeRepetitionsAll groups could do?All individuals could do?
semantic decisionButton press5NoNo
tone decisionButton press6UnknownUnknown
Conditions notesGroup performance below chance; several patients at 0 which is difficult to understand in a 2AFC task

Contrasts

Are the contrasts clearly described?Yes

Contrast 1: semantic decision vs tone decision

Language conditionSemantic decision
Control conditionTone decision
Are the conditions matched for visual demands?Yes
Are the conditions matched for auditory demands?Yes
Are the conditions matched for motor demands?Yes
Are the conditions matched for cognitive/executive demands?Yes
Is accuracy matched between the language and control tasks for all relevant groups?Unknown, not reported
Is reaction time matched between the language and control tasks for all relevant groups?Unknown, not reported
Behavioral data notesTone decision accuracy not reported
Are control data reported in this paper or another that is referenced?Yes
Does the contrast selectively activate plausible relevant language regions in the control group?Yes
Are activations lateralized in the control data?Yes
Control activation notesTemporal activation is mid MTG and AG rather than pSTS
Contrast notes

Analyses

Are the analyses clearly described?Yes

ROI analysis 1

First level contrastSemantic decision vs tone decision
Analysis classCross-sectional correlation with language or other measure
Group(s)Aphasia
CovariateSemantic decision accuracy
Is the second level contrast valid in terms of the group(s), time point(s), and measures involved?Yes
Is accuracy matched across the second level contrast?Accuracy is covariate
Is reaction time matched across the second level contrast?Unknown, not reported
Behavioral data notes
Type of analysisRegions of interest (ROI)
ROI typeOther
How many ROIs are there?3
What are the ROI(s)?(1) L AG and bilateral midline components of the canonical semantic network, along with reduced activity in R frontal, temporal and parietal regions; (2) bilateral IFG pars orbitalis; (3) L IFG and DLPFC along with bilateral midline regions
How are the ROI(s) defined?ROIs are mixing coefficients of functional networks arising from mCCA + jICA that were differently represented in the patient and control groups
Correction for multiple comparisonsFamilywise error (FWE)
Statistical details
Findings↑ L IFG
↑ L dorsolateral prefrontal cortex
↑ L SMA/medial prefrontal
↑ L angular gyrus
↑ L precuneus
↑ L posterior cingulate
↑ R IFG pars orbitalis
↑ R SMA/medial prefrontal
↑ R precuneus
↑ R posterior cingulate
↓ L insula
↓ R IFG pars opercularis
↓ R IFG pars triangularis
↓ R insula
↓ R dorsal precentral
↓ R supramarginal gyrus
↓ R posterior STG
↓ R mid temporal
Findings notesAll 3 networks were significantly correlated; analysis of networks so involvement of each individual region cannot be assured

ROI analysis 2

First level contrastSemantic decision vs tone decision
Analysis classCross-sectional correlation with language or other measure
Group(s)Aphasia
CovariateAverage of semantic and phonemic fluency
Is the second level contrast valid in terms of the group(s), time point(s), and measures involved?Yes
Is accuracy matched across the second level contrast?Unknown, not reported
Is reaction time matched across the second level contrast?Unknown, not reported
Behavioral data notes
Type of analysisRegions of interest (ROI)
ROI typeOther
How many ROIs are there?3
What are the ROI(s)?(1) L AG and bilateral midline components of the canonical semantic network, along with reduced activity in R frontal, temporal and parietal regions; (2) bilateral IFG pars orbitalis; (3) L IFG and DLPFC along with bilateral midline regions
How are the ROI(s) defined?ROIs are mixing coefficients of functional networks arising from mCCA + jICA that were differently represented in the patient and control groups
Correction for multiple comparisonsFamilywise error (FWE)
Statistical details
Findings↑ L IFG
↑ L dorsolateral prefrontal cortex
↑ L SMA/medial prefrontal
↑ L angular gyrus
↑ L precuneus
↑ L posterior cingulate
↑ R SMA/medial prefrontal
↑ R precuneus
↑ R posterior cingulate
↓ L insula
↓ R IFG pars opercularis
↓ R IFG pars triangularis
↓ R insula
↓ R dorsal precentral
↓ R supramarginal gyrus
↓ R posterior STG
↓ R mid temporal
Findings notesNetworks 1 and 3 were significantly correlated; analysis of networks so involvement of each individual region cannot be assured

ROI analysis 3

First level contrastSemantic decision vs tone decision
Analysis classCross-sectional correlation with language or other measure
Group(s)Aphasia
CovariateBNT
Is the second level contrast valid in terms of the group(s), time point(s), and measures involved?Yes
Is accuracy matched across the second level contrast?Unknown, not reported
Is reaction time matched across the second level contrast?Unknown, not reported
Behavioral data notes
Type of analysisRegions of interest (ROI)
ROI typeOther
How many ROIs are there?3
What are the ROI(s)?(1) L AG and bilateral midline components of the canonical semantic network, along with reduced activity in R frontal, temporal and parietal regions; (2) bilateral IFG pars orbitalis; (3) L IFG and DLPFC along with bilateral midline regions
How are the ROI(s) defined?ROIs are mixing coefficients of functional networks arising from mCCA + jICA that were differently represented in the patient and control groups
Correction for multiple comparisonsFamilywise error (FWE)
Statistical details
Findings↑ L IFG
↑ L dorsolateral prefrontal cortex
↑ L SMA/medial prefrontal
↑ L angular gyrus
↑ L precuneus
↑ L posterior cingulate
↑ R SMA/medial prefrontal
↑ R precuneus
↑ R posterior cingulate
↓ L insula
↓ R IFG pars opercularis
↓ R IFG pars triangularis
↓ R insula
↓ R dorsal precentral
↓ R supramarginal gyrus
↓ R posterior STG
↓ R mid temporal
Findings notesNetworks 1 and 3 were significantly correlated; analysis of networks so involvement of each individual region cannot be assured

Complex analysis 1

First level contrastSemantic decision vs tone decision
Analysis classCross-sectional aphasia vs control
Group(s)Aphasia vs control
Covariate
Is the second level contrast valid in terms of the group(s), time point(s), and measures involved?Yes
Is accuracy matched across the second level contrast?No, different
Is reaction time matched across the second level contrast?Unknown, not reported
Behavioral data notesSemantic decision accuracy not matched, but tone decision accuracy not reported
Type of analysisComplex
Statistical detailsMultimodal canonical correlation analysis (mCCA) and joint ICA were used to identify 3 joint ICs (structural/functional) that were differently represented in the patient and control groups. Although there was no correction for multiple comparisons when the functional maps were thresholded, the maps for the three networks each appeared to relate to coherent parts of the semantic network.
FindingsOther
Findings notesThe first joint IC comprised preservation of tissue in L posterior temporo-parietal region, activity in the L AG and bilateral midline components of the canonical semantic network, and reduced activity in R frontal, temporal and parietal regions. The second joint IC comprised preservation of tissue in the the L basal ganglia/insula region, and activity predominantly in the IFG pars orbitalis bilaterally. The third joint IC comprised preservation of the L IFG and activity in the L IFG and DLPFC along with bilateral midline regions. The first joint IC was considered to provide more robust evidence for structure-function relationships than the other two, because it was the only one where individual structural and functional mixing coefficients remained correlated even when lesion volume was included as a covariate.

Notes

Excluded analyses(1) group analyses that were described in a previous paper (Griffis et al., 2017, Hum Brain Mapp); (2) ancillary analysis using different numbers of components per modality; (3) ancillary analysis using lesion masks instead of brain tissue maps; (4) ancillary analysis using multivariate lesion-symptom mapping, because these analyses yielded similar results to the main analysis