White matter structural connectivity is associated with sensorimotor function in stroke survivors

Abstract

PurposeDiffusion tensor imaging (DTI) provides functionally relevant information about white matter structure. Local anatomical connectivity information combined with fractional anisotropy (FA) and mean diffusivity (MD) may predict functional outcomes in stroke survivors. Imaging methods for predicting functional outcomes in stroke survivors are not well established. This work uses DTI to objectively assess the effects of a stroke lesion on white matter structure and sensorimotor function.MethodsA voxel-based approach is introduced to assess a stroke lesion's global impact on motor function. Anatomical T1-weighted and diffusion tensor images of the brain were acquired for nineteen subjects (10 post-stroke and 9 age-matched controls). A manually selected volume of interest was used to alleviate the effects of stroke lesions on image registration. Images from all subjects were registered to the images of the control subject that was anatomically closest to Talairach space. Each subject's transformed image was uniformly seeded for DTI tractography. Each seed was inversely transformed into the individual subject space, where DTI tractography was conducted and then the results were transformed back to the reference space. A voxel-wise connectivity matrix was constructed from the fibers, which was then used to calculate the number of directly and indirectly connected neighbors of each voxel. A novel voxel-wise indirect structural connectivity (VISC) index was computed as the average number of direct connections to a voxel's indirect neighbors. Voxel-based analyses (VBA) were performed to compare VISC, FA, and MD for the detection of lesion-induced changes in sensorimotor function. For each voxel, a t-value was computed from the differences between each stroke brain and the 9 controls. A series of linear regressions was performed between Fugl-Meyer (FM) assessment scores of sensorimotor impairment and each DTI metric's log number of voxels that differed from the control group.ResultsCorrelation between the logarithm of the number of significant voxels in the ipsilesional hemisphere and total Fugl-Meyer score was moderate for MD (R2 = 0.512), and greater for VISC (R2 = 0.796) and FA (R2 = 0.674). The slopes of FA (p = 0.0036), VISC (p = 0.0005), and MD (p = 0.0199) versus the total FM score were significant. However, these correlations were driven by the upper extremity motor component of the FM score (VISC: R2 = 0.879) with little influence of the lower extremity motor component (FA: R2 = 0.177).ConclusionThe results suggest that a voxel-wise metric based on DTI tractography can predict upper extremity sensorimotor function of stroke survivors, and that supraspinal intraconnectivity may have a less dominant role in lower extremity function.

Age-related Structural and Functional Variations in 5967 Individuals across the Adult Lifespan

“…Several previous studies reported that regions with few or no direct structural connections may exhibit high functional connectivity, which indicates presence of indirect connections between structure and function (Damoiseaux & Greicius, 2009;Honey et al, 2009). Kalinosky et al further reported that the structural connectivity changes constructed using dMRI data were not related to the corresponding changes in FNC within RSNs (Kalinosky, Schindler-Ivens, & Schmit, 2013). In this study, we observed two matched SNC and FNC cells exhibiting a similar age-related changing trend across the adult lifespan as shown in Figure 5: 1) the connectivity between middle occipital gyrus and insula; 2) the connectivity between precuneus and cerebellum.…”

Structural connectivity analyses in motor recovery research after stroke

“…The references are numbered consecutively and listed below with the first author and the year of publication, et al has been omitted for sake of readability. 1: Werring 2000; 2: Pierpaoli 2001; 3: Thomalla 2004; 4: Konishi 2005; 5: Newton 2006; 6: Gupta 2006; 7: Liang 2007; 8: Stinear 2007; 9: Schaechter 2008; 10: Jang 2008; 11: Nelles 2008; 12: Schaechter 2009; 13: Sterr 2010; 14: Radlinska 2010; 15: Lindenberg 2010; 16: Yeo and Jang 2010; 17: Zhu 2010; 18: Puig 2010; 19: Crofts 2011; 20: Bosnell 2011; 21: Qiu 2011; 22: Riley 2011; 23: Kwon 2011; 24: Puig 2011; 25: Nouri 2011; 26: Granziera 2012; 27: Borich 2012; 28: Schulz 2012; 29: Lotze 2012; 30: Lindenberg 2012; 31: Wang 2012; 32: Rüber 2012; 33: Radlinska 2012; 34: Carter 2012; 35: Stinear 2012; 36: Kalinosky 2013; 37: Chen and Schlaug 2013; 38: Vargas 2013; 39: Park 2013; 40: Kou 2013; 41: Phan 2013; 42: Puig 2013; 43: Groisser 2014; 44: Takenobu 2014; 45: Sterr 2014; 46: Kuceyeski 2014; 47: Lin 2015; 48: Schulz 2015; 49: Volz 2015; 50: Li 2015; 51: Song 2015; 52: Zheng and Schlaug 2015; 53: Liu 2015; 54: Schulz 2015; 55: Feng 2015; 56: Byblow 2015 …”

Structural and Resting-State Brain Connectivity of Motor Networks After Stroke

“…27 Although these tractography studies may be informative, they do require tracing of individual tracts between seed regions and may thus not yet be applicable for routine clinical use. Approaches that try to automatically map all possible fiber tract connections (the connectome) using voxelwise connectivity matrices 28 or try to estimate remote stroke effects from normal connectivity maps 29 may be promising alternatives to tractography, but require further exploration.…”