Automatic identification of brain lesions from magnetic resonance imaging (MRI) scans of stroke survivors would be a useful aid in patient diagnosis and treatment planning. It would also greatly facilitate the study of brain-behavior relationships by eliminating the laborious step of having a human expert manually segment the lesion on each brain scan. We propose a multi-modal multi-path convolutional neural network system for automating stroke lesion segmentation. Our system has nine end-to-end UNets that take as input 2-dimensional (2D) slices and examines all three planes with three different normalizations. Outputs from these nine total paths are concatenated into a 3D volume that is then passed to a 3D convolutional neural network to output a final lesion mask. We trained and tested our method on datasets from three sources: Medical College of Wisconsin (MCW), Kessler Foundation (KF), and the publicly available Anatomical Tracings of Lesions After Stroke (ATLAS) dataset. To promote wide applicability, lesions were included from both subacute (1 to 5 weeks) and chronic ( > 3 months) phases post stroke, and were of both hemorrhagic and ischemic etiology. Cross-study validation results (with independent training and validation datasets) were obtained to compare with previous methods based on naive Bayes, random forests, and three recently published convolutional neural networks. Model performance was quantified in terms of the Dice coefficient, a measure of spatial overlap between the model-identified lesion and the human expert-identified lesion, where 0 is no overlap and 1 is complete overlap. Training on the KF and MCW images and testing on the ATLAS images yielded a mean Dice coefficient of 0.54. This was reliably better than the next best previous model, UNet, at 0.47. Reversing the train and test datasets yields a mean Dice of 0.47 on KF and MCW images, whereas the next best UNet reaches 0.45. With all three datasets combined, the current system compared to previous methods also attained a reliably higher cross-validation accuracy. It also achieved high Dice values for many smaller lesions that existing methods have difficulty identifying. Overall, our system is a clear improvement over previous methods for automating stroke lesion segmentation, bringing us an important step closer to the inter-rater accuracy level of human experts.

Emerging research suggests spatial neglect after right stroke is linked to dysfunctional attention and motor networks. Advanced functional connectivity analysis clarified brain network recovery, however we need to know how networks participate in adaptive motor performance. We need to verify network changes associated with validated functional measures and spatial-motor performance in spatial neglect, especially in patients with large brain lesions and significant disability. This study tested whether disability-relevant spatial neglect associates with different patterns of resting state functional connectivity between motor, dorsal and ventral attention networks (MN, DAN and VAN). Right stroke patients had spatial neglect (n = 8) or not (n = 10) on the Behavioural Inattention Test-conventional. Spatial neglect patients had weaker intranetwork VAN connectivity, and reduced internetwork connectivity between VAN and left frontal eye field (DAN), and between VAN and the left primary motor area (MN). These network impairments might explain the co-occurrence of attention and motor deficits in spatial neglect, and open a path to assessing functional connectivity in clinical trials of combined spatial retraining and motor rehabilitation after stroke.

Functional disability due to spatial neglect hinders recovery in up to 30% of stroke survivors. Prism adaptation treatment (PAT) may alleviate the disabling consequences of spatial neglect, but we do not yet know why some individuals show much better outcomes following PAT than others. The goal of this scoping review and meta-analysis was to investigate the neural mechanisms underlying prism adaptation (PA). We conducted both quantitative and qualitative analyses across fMRI studies investigating brain activity before, during, and after PA, in healthy individuals and patients with right or left brain damage (RBD or LBD) due to stroke. In healthy adults, PA was linked with activity in posterior parietal and cerebellar clusters, reduced bilateral parieto-frontal connectivity, and increased fronto-limbic and sensorimotor network connectivity. In contrast, RBD individuals with spatial neglect relied on different circuits, including an activity cluster in the intact left occipital cortex. This finding is consistent with a shift in hemispheric dominance in spatial processing to the left hemisphere. However, more studies are needed to clarify the contribution of lesion location and load on the circuits involved in PA after unilateral brain damage. Future studies are also needed to clarify the relationship of decreasing resting state functional connectivity (rsFC) to visuomotor function.

Background and Purpose: Delirium, an acute reduction in cognitive functioning, hinders stroke recovery and contributes to cognitive decline. Right-hemisphere stroke is linked with higher delirium incidence, likely, due to the prevalence of spatial neglect (SN), a right-brain disorder of spatial processing. This study tested if symptoms of delirium and SN after right-hemisphere stroke are associated with abnormal function of the right-dominant neural networks specialized for maintaining attention, orientation, and arousal. Methods: Twenty-nine participants with right-hemisphere ischemic stroke undergoing acute rehabilitation completed delirium and SN assessments and functional neuroimaging scans. Whole-brain functional connectivity of 4 right-hemisphere seed regions in the cortical-subcortical arousal and attention networks was assessed for its relationship to validated SN and delirium severity measures. Results: Of 29 patients, 6 (21%) met the diagnostic criteria for delirium and 16 (55%) for SN. Decreased connectivity of the right basal forebrain to brain stem and basal ganglia predicted more severe SN. Increased connectivity of the arousal and attention network regions with the parietal, frontal, and temporal structures in the unaffected hemisphere was also found in more severe delirium and SN. Conclusions: Delirium and SN are associated with decreased arousal network activity and an imbalance of cortico-subcortical hemispheric connectivity. Better understanding of neural correlates of poststroke delirium and SN will lead to improved neuroscience-based treatment development for these disorders. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT03349411.

Background and Purpose: Delirium portends worse outcomes after intracerebral hemorrhage (ICH), but it is unclear if symptom resolution or postacute care intensity may mitigate its impact. We aimed to explore differences in outcome associated with delirium resolution before hospital discharge, as well as the potential mediating role of postacute discharge site. Methods: We performed a single-center cohort study on consecutive ICH patients over 2 years. Delirium was diagnosed according to DSM-5 criteria and further classified as persistent or resolved based on delirium status at hospital discharge. We determined the impact of delirium on unfavorable 3-month outcome (modified Rankin Scale score, 4–6) using logistic regression models adjusted for established ICH predictors, then used mediation analysis to examine the indirect effect of delirium via postacute discharge site. Results: Of 590 patients (mean age 70.5±15.5 years, 52% male, 83% White), 59% (n=348) developed delirium during hospitalization. Older age and higher ICH severity were delirium risk factors, but only younger age predicted delirium resolution, which occurred in 75% (161/215) of ICH survivors who had delirium. Delirium was strongly associated with unfavorable outcome, but patients with persistent delirium fared worse (adjusted odds ratio [OR], 7.3 [95% CI, 3.3–16.3]) than those whose delirium resolved (adjusted OR, 3.1 [95% CI, 1.8–5.5]). Patients with delirium were less likely to be discharged to inpatient rehabilitation than skilled nursing facilities (adjusted OR, 0.31 [95% CI, 0.17–0.59]), and postacute care site partially mediated the relationship between delirium and functional outcome in ICH survivors, leading to a 25% reduction in the effect of delirium (without mediator: adjusted OR, 3.0 [95% CI, 1.7–5.6]; with mediator: adjusted OR, 2.3 [95% CI, 1.2–4.3]). Conclusions: Acute delirium resolves in most patients with ICH by hospital discharge, which was associated with better outcomes than in patients with persistent delirium. The impact of delirium on outcomes may be further mitigated by postacute rehabilitation.

Better understanding of cerebral blood flow (CBF) perfusion in stroke recovery can help inform decisions about optimal timing and targets of restorative treatments. In this study, we examined the relationship between cerebral perfusion and recovery from stroke‐induced reading deficits. Left stroke patients were tested with a noninvasive CBF measure (arterial spin labeling) 3 months post‐stroke. We measured blood flow perfusion within the left and right sides of the brain, in areas surrounding the lesion, and areas belonging to the reading network. Two hypotheses were tested. The first was that recovery of reading function depends on increased perfusion around the stroke lesion. This hypothesis was not supported by our findings. The second hypothesis was that increased perfusion of intact areas within the reading circuit is tightly coupled with recovery. Our findings are consistent with this hypothesis. Specifically, higher perfusion in the left reading network measured during the subacute stroke period predicted better reading ability and phonology competence in the chronic period. In contrast, higher perfusion of the right homologous regions was associated with decreased reading accuracy and phonology competence in the subacute and chronic periods. These findings suggest that recovery of reading and language competence may rely on improved blood flow in the reading network of the language‐dominant hemisphere.