Connection between microstructural alterations detected by diffusion MRI and cognitive dysfunction in MS: A model-free analysis approach

Abstract

Background: Cognitive decline is a prominent symptom of MS. Clear connection between cognitive status and white matter microstructural changes has not been unequivocally observed to date.

Objective: To characterise the relationship between white matter microstructure and cognitive performance a partial least squares (PLS) approach was used.

Methods: 53 RR MS patients' T1 and DTI images and BICAMS subtests were used in our analysis. Standard FSL pipeline was used to obtain diffusion parameters. A PLS approach was applied to reveal the diffusion parameter patterns responsible for the cognitive dysfunction.

Results: The first latent variable (LV) was mainly associated with demyelination, while the second and third explained axonal damage. While the first two LV represented mainly Brief Visuospatial Memory Test (BVMT) and Single Digit Modality Test (SDMT), the third LV depicted diffusion alterations mainly the verbal subtest. The first LVs spatial map showed demyelination in the corpus callosum. The second LVs spatial map showed the diffusion alterations in the thalamus. The third LV depicted diffusion alterations in the putative left superior longitudinal fascicle.

Conclusion: Visual memory demanding tasks versus language functions depend on distinct patterns of diffusion parameters and the spatial organisation. Axial diffusivity alterations, a putative marker of irreversible axonal loss explained around 20% of variability in the cognitive functions.

Keywords: BICAMS; DTI; PLS; TBSS.