Highlights d White matter structural disconnections explain brain network dysfunction after stroke d Damage to the gray matter, including ''hub'' regions, provides less explanatory power d Interhemispheric disconnections are linked to widespread functional disruptions d Disconnection and disruption are topographically linked within functional networks

Damage to the white matter underlying the left posterior temporal lobe leads to deficits in multiple language functions. The posterior temporal white matter may correspond to a bottleneck where both dorsal and ventral language pathways are vulnerable to simultaneous damage. Damage to a second putative white matter bottleneck in the left deep prefrontal white matter involving projections associated with ventral language pathways and thalamo-cortical projections has recently been proposed as a source of semantic deficits after stroke. Here, we first used white matter atlases to identify the previously described white matter bottlenecks in the posterior temporal and deep prefrontal white matter. We then assessed the effects of damage to each region on measures of verbal fluency, picture naming, and auditory semantic decision-making in 43 chronic left hemispheric stroke patients. Damage to the posterior temporal bottleneck predicted deficits on all tasks, while damage to the anterior bottleneck only significantly predicted deficits in verbal fluency. Importantly, the effects of damage to the bottleneck regions were not attributable to lesion volume, lesion loads on the tracts traversing the bottlenecks, or damage to nearby cortical language areas. Multivariate lesion-symptom mapping revealed additional lesion predictors of deficits. Post-hoc fiber tracking of the peak white matter lesion predictors using a publicly available tractography atlas revealed evidence consistent with the results of the bottleneck analyses. Together, our results provide support for the proposal that spatially specific white matter damage affecting bottleneck regions, particularly in the posterior temporal lobe, contributes to chronic language deficits after left hemispheric stroke. This may reflect the simultaneous disruption of signaling in dorsal and ventral language processing streams.

Background Manual lesion delineation by an expert is the standard for lesion identification in MRI scans, but is time-consuming and can introduce subjective bias. Alternative methods often require multi-modal MRI data, user interaction, scans from a control population, and/or arbitrary statistical thresholding. New Method We present an approach for automatically identifying stroke lesions in individual T1-weighted MRI scans using naïve Bayes classification. Probabilistic tissue segmentation and image algebra were used to create feature maps encoding information about missing and abnormal tissue. Leave-one-case-out training and cross-validation was used to obtain out-of-sample predictions for each of 30 cases with left hemisphere stroke lesions. Results Our method correctly predicted lesion locations for 30/30 un-trained cases. Post-processing with smoothing (8mm FWHM) and cluster-extent thresholding (100 voxels) was found to improve performance. Comparison with Existing Method Quantitative evaluations of post-processed out-of-sample predictions on 30 cases revealed high spatial overlap (mean Dice similarity coefficient = 0.66) and volume agreement (mean percent volume difference = 28.91; Pearson’s r = 0.97) with manual lesion delineations. Conclusions Our automated approach agrees with manual tracing. It provides an alternative to automated methods that require multi-modal MRI data, additional control scans, or user interaction to achieve optimal performance. Our fully trained classifier has applications in neuroimaging and clinical contexts.

No abstract

Functional connectivity (FC) studies have identified physiological signatures of stroke that correlate with behavior. Using structural and functional MRI data from 114 stroke patients, 24 matched controls, and the Human Connectome Project, we tested the hypothesis that structural disconnection, not damage to critical regions, underlies FC disruptions. Disconnection severity outperformed damage to putative FC connector nodes for explaining reductions in system modularity, and multivariate models based on disconnection outperformed damage models for explaining FC disruptions within and between systems. Across patients, disconnection and FC patterns exhibited a lowdimensional covariance dominated by a single axis linking interhemispheric disconnections to reductions in FC measures of interhemispheric system integration, ipsilesional system segregation, and system modularity, and that correlated with multiple behavioral deficits. These findings clarify the structural basis of FC disruptions in stroke patients and demonstrate a low-dimensional link between perturbations of the structural connectome, disruptions of the functional connectome, and behavioral deficits. Stroke disrupts system-scale functional connectivityResting-state fMRI data were used to measure FC between 324 cortical parcels associated with different brain systems ( Fig. 2A-C). We defined twelve system-scale summary measures to capture reductions of interhemispheric system integration, ipsilesional system segregation, and system modularity. For each patient, we extracted the mean interhemispheric FC values for nine bilateral cortical systems (Fig. 2D, left) and averaged across systems to summarize interhemispheric system integration across the cortex (Fig. 2D, left inset). We also extracted the mean FC between the ipsilesional DAN and DMN to quantify ipsilesional system segregation (Fig. 2D, middle), and we averaged modularity estimates for a priori system partitions across multiple edge density thresholds to summarize overall network structure ( Fig. 2D, right; mean shown in inset).The mean FC matrices for patients and controls had similar topographies ( Fig. 2B; r=0.96, p<0.001), but subtracting the patient matrix from the control matrix revealed magnitude differences that were often in opposite directions for connections with positive vs. negative values in the mean control matrix (Fig. 2C; r=-0.42, p<0.001), consistent with reduced integration within systems and reduced segregation between systems in patients. As expected, patients showed marked abnormalities in FC summary measures of interhemispheric integration, ipsilesional segregation, and system modularity (Fig. 2D). These measures were used as dependent variables in subsequent analyses.

Highlights Freely available MATLAB toolkit for creating anatomically informed lesion measures. Estimates region-level grey matter damage and white matter disconnections. Estimates tract-level disconnections and produces disconnection maps. Estimates increases in region-level shortest path lengths associated with the lesion. Can produce region-level measures for user-selected brain parcellations.

No abstract

Current theories of language recovery after stroke are limited by a reliance on small studies. Here, we aimed to test predictions of current theory and resolve inconsistencies regarding right hemispheric contributions to long-term recovery. We first defined the canonical semantic network in 43 healthy controls. Then, in a group of 43 patients with chronic post-stroke aphasia, we tested whether activity in this network predicted performance on measures of semantic comprehension, naming, and fluency while controlling for lesion volume effects. Canonical network activation accounted for 22–33% of the variance in language test scores. Whole-brain analyses corroborated these findings, and revealed a core set of regions showing positive relationships to all language measures. We next evaluated the relationship between activation magnitudes in left and right hemispheric portions of the network, and characterized how right hemispheric activation related to the extent of left hemispheric damage. Activation magnitudes in each hemispheric network were strongly correlated, but four right frontal regions showed heightened activity in patients with large lesions. Activity in two of these regions (inferior frontal gyrus pars opercularis and supplementary motor area) was associated with better language abilities in patients with larger lesions, but poorer language abilities in patients with smaller lesions. Our results indicate that bilateral language networks support language processing after stroke, and that right hemispheric activations related to extensive left hemisphere damage occur outside of the canonical semantic network and differentially relate to behavior depending on the extent of left hemispheric damage.