Premium
A novel MRI contrast weighted ratio method for measuring myelin in older adults at risk for Alzheimer’s disease
Author(s) -
Price Lauren Rose,
Deardorff Rachael,
Wu YuChien,
West John D.,
McDonald Brenna C.,
Risacher Shan L.,
Saykin Andrew J.
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.046297
Subject(s) - fluid attenuated inversion recovery , fornix , hyperintensity , white matter , bonferroni correction , psychology , cognitive decline , cognition , audiology , magnetic resonance imaging , neuroimaging , effects of sleep deprivation on cognitive performance , cognitive impairment , medicine , nuclear medicine , pathology , neuroscience , disease , dementia , radiology , mathematics , hippocampus , statistics
Background MRI measures sensitive to myelin integrity such as the T1/T2 weighted ratio method (T1w/T2w; Glasser & Van Essen 2011) have been investigated as potential AD biomarkers. Many cohorts have transitioned to T1 and T2‐weighted Fluid Attenuated Inversion Recovery (FLAIR) acquisitions. Our goal was to validate the T1w/T2FLAIRw method in comparison to the T1w/T2w method in the same cohort. Method 28 cognitively unimpaired (CU, cognitively normal and subjective cognitive decline) and 26 AD‐related cognitively impaired (CI, mild cognitive impairment and AD) individuals were analyzed to determine the correlative validity of association between T1w/T2FLAIRw and T1w/T2w methods. Image processing algorithms were designed in‐house for application to AD, building on previous methods and implemented in SPM12. White matter regions of interest (ROIs) from the Johns Hopkins ICBM‐DTI‐81 white‐matter labels atlas were masked excluding cerebrospinal fluid and white matter hyperintensities. Pearson correlation was used to test associations between the canonical method and our extension. Then, diagnostic group differences were assessed using regional independent sample T‐tests, with Bonferroni correction for independent components (8 independent factors identified by PCA of all white matter tracts). Result Myelinated regions associated with cognitive decline, as well as all other atlas ROIs, demonstrated significant (p <.01) positive correlations between T1w/T2w and T1w/T2FLAIR values (r values between .590 and .915) with the exception of the fornix, which was significant at the .05 level in CU. Between‐group analyses using T1w/T2FLAIRw showed lower ratios (reflecting reduced myelin) in CI. Significant differences in group T1w/T2FLAIRw ratios were seen in the left hippocampal cingulum (left, p=.011; right, p=.024), uncinate fasciculus (left, p=.007; right, p=.015), and the right anterior limb of the internal capsule (p=.019) and superior fronto‐occipital fasciculus (p=.011). After Bonferroni correction only the left hippocampal cingulum, left uncinate fasciculus and right superior fronto‐occipital fasciculus approached significance (p = .088; .088; .056 respectively). Conclusion Results suggest T1w/T2FLAIRw is a valid alternative to T1w/T2w ratios. Between groups, T1w/T2FLAIRw pointed towards lower myelin in intersecting white matter tracts in cognitively impaired individuals. Although initial results appear promising with reasonably large effect sizes, a larger sample needs to be assessed to provide sufficient statistical power to effectively assess the proposed method.