Enhanced estimations of post-stroke aphasia severity using stacked multimodal predictions.

The severity of post-stroke aphasia and the potential for recovery are highly variable and difficult to predict. Evidence suggests that optimal estimation of aphasia severity requires the integration of multiple neuroimaging modalities and the adoption of new methods that can detect multivariate brain-behavior relationships. We created and tested a multimodal framework that relies on three information sources (lesion maps, structural connectivity, and functional connectivity) to create an array of unimodal predictions which are then fed into a final model that creates "stacked multimodal predictions" (STAMP). Crossvalidated predictions of four aphasia scores (picture naming, sentence repetition, sentence comprehension, and overall aphasia severity) were obtained from 53 left hemispheric chronic stroke patients (age: 57.1 ± 12.3 yrs, post-stroke interval: 20 months, 25 female). Results showed accurate predictions for all four aphasia scores (correlation true vs. predicted: r = 0.79-0.88). The accuracy was slightly smaller but yet significant (r = 0.66) in a full split crossvalidation with each patient considered as new. Critically, multimodal predictions produced more accurate results that any single modality alone. Topological maps of the brain regions involved in the prediction were recovered and compared with traditional voxel-based lesion-to-symptom maps, revealing high spatial congruency. These results suggest that neuroimaging modalities carry complementary information potentially useful for the prediction of aphasia scores. More broadly, this study shows that the translation of neuroimaging findings into clinically useful tools calls for a shift in perspective from unimodal to multimodal neuroimaging, from univariate to multivariate methods, from linear to nonlinear models, and, conceptually, from inferential to predictive brain mapping. Hum Brain Mapp 38:5603-5615, 2017. © 2017 Wiley Periodicals, Inc.

Critical brain regions for action recognition: lesion symptom mapping in left hemisphere stroke.

A number of conflicting claims have been advanced regarding the role of the left inferior frontal gyrus, inferior parietal lobe and posterior middle temporal gyrus in action recognition, driven in part by an ongoing debate about the capacities of putative mirror systems that match observed and planned actions. We report data from 43 left hemisphere stroke patients in two action recognition tasks in which they heard and saw an action word ('hammering') and selected from two videoclips the one corresponding to the word. In the spatial recognition task, foils contained errors of body posture or movement amplitude/timing. In the semantic recognition task, foils were semantically related (sawing). Participants also performed a comprehension control task requiring matching of the same verbs to objects (hammer). Using regression analyses controlling for both the comprehension control task and lesion volume, we demonstrated that performance in the semantic gesture recognition task was predicted by per cent damage to the posterior temporal lobe, whereas the spatial gesture recognition task was predicted by per cent damage to the inferior parietal lobule. A whole-brain voxel-based lesion symptom-mapping analysis suggested that the semantic and spatial gesture recognition tasks were associated with lesioned voxels in the posterior middle temporal gyrus and inferior parietal lobule, respectively. The posterior middle temporal gyrus appears to serve as a central node in the association of actions and meanings. The inferior parietal lobule, held to be a homologue of the monkey parietal mirror neuron system, is critical for encoding object-related postures and movements, a relatively circumscribed aspect of gesture recognition. The inferior frontal gyrus, on the other hand, was not predictive of performance in any task, suggesting that previous claims regarding its role in action recognition may require refinement.