Structural connectome disruption at baseline predicts 6-months post-stroke outcome

Abstract

In this study, models based on quantitative imaging biomarkers of post-stroke structural connectome disruption were used to predict six-month outcomes in various domains. Demographic information and clinical MRIs were collected from 40 ischemic stroke subjects (age: 68.1±13.2 years, 17 female, NIHSS: 6.8±5.6). Diffusion-weighted images were used to create lesion masks, which were uploaded to the Network Modification (NeMo) Tool. The NeMo Tool, using only clinical MRIs, allows estimation of connectome disruption at three levels: whole brain, individual gray matter regions and between pairs of gray matter regions. Partial Least Squares Regression models were constructed for each level of connectome disruption and for each of the three six-month outcomes: applied cognitive, basic mobility and daily activity. Models based on lesion volume were created for comparison. Cross-validation, bootstrapping and multiple comparisons corrections were implemented to minimize over-fitting and Type I errors. The regional disconnection model best predicted applied cognitive (R2 = 0.56) and basic mobility outcomes (R2 = 0.70), while the pairwise disconnection model best predicted the daily activity measure (R2 = 0.72). These results demonstrate that models based on connectome disruption metrics were more accurate than ones based on lesion volume and that increasing anatomical specificity of disconnection metrics does not always increase model accuracy, likely due to statistical adjustments for concomitant increases in data dimensionality. This work establishes that the NeMo Tool's measures of baseline connectome disruption, acquired using only routinely collected MRI scans, can predict 6-month post-stroke outcomes in various functional domains including cognition, motor function and daily activities.

Enhanced estimations of post‐stroke aphasia severity using stacked multimodal predictions

“…However, disconnections can occur at any segment along WM tracts, and, consequently, lesions apparently unrelated to each other can cause disruption of the same tract, producing similar symptoms. Supporting this rationale, recent studies have shown that structural disconnections can predict cognitive performance more accurately than lesion location [Hope et al, ; Kuceyeski et al, ]. Nonoverlapping lesions can also affect the same functional network, leading to similar clinical syndromes [Boes et al, ].…”

“…More recently, Yourganov et al [] created predictive models of several aphasia scores from 90 subjects using either lesion or structural connectivity information, and achieved correlation scores in the range r = 0.52–0.72. In another study, Kuceyeski et al [] predicted future cognitive abilities in post‐stroke subjects using virtual tractography lesions, achieving an accuracy of r = 0.75. Compared to the above literature, STAMP showed higher predicted vs. true correlation ( r = 0.79–0.88) by integrating multimodal information.…”

“…A large sample size may also increase the chances that similar brain architectures exist in the training and test groups [Price et al, ]. It is important to note that all previous studies have performed variable selection on the whole group [Hope et al, ; Kuceyeski et al, ]. An exception to this trend is the study of Yourganov et al [], who selected the variables only from the training group.…”

“…The importance of lesion location as a determinant of post‐stroke aphasia is also reflected in translational models that aim to predict aphasia severity [Basilakos et al, ; Hope et al, ; Seghier et al, ; Wang et al, ; Yourganov et al, ]. However, there is evidence that lesion location provides only a partial picture of the factors that contribute to cognitive performance after stroke [Carter et al, ; Hope et al, ; Kuceyeski et al, ; Willmes and Poeck, ]. This is likely because performance does not depend solely on the “missing” parts of the brain, but also on the structural and functional integrity of the remaining tissue, which can support the functional reorganization of language processes [Carter et al, ; Corbetta et al, ; Crofts et al, ; Rorden and Karnath, ; Saur et al, ; Siegel et al, ; Turkeltaub et al, ; Wu et al, ].…”

Neuropsychology and cognitive neuroscience in the fMRI era: A recapitulation of localizationist and connectionist views.

“…A traditional lesion overlap analysis showed these patients had largely non-overlapping damage; however, lesion-network mapping demonstrated that the patients who were impaired on the IGT showed overlap in their lesion connectivity maps with somatosensory, motor and insula cortices, to a greater extent than patients who had unimpaired performance on the IGT. An analogous approach utilizing diffusion tractography imaging (DTI) data has also shown promise in understanding outcome from stroke (Kuceyeski et al, 2016). Future refinements to these network-derived lesion-mapping approaches (for example, combining DTI and resting-state data) hold further promise for better appreciating the remote effects of damage in the context of neuropsychological studies of brain and behavior.…”