Evaluation of Common Changes in the Aging Brain: Comparing High- and Low-Field MRI

Xiaowei Song,Hui Guo,Yunting Zhang,Kenneth Rockwood
DOI: https://doi.org/10.1016/j.jalz.2013.05.193
2013-01-01
Abstract:Multiple brain structural changes occur during aging, including atrophy, white matter lesions, and small vessel damage. These changes typically are more common and more severe in Alzheimer's disease (AD). Recently, a Brain Atrophy and Lesion Index (BALI) has been established using high-field MRI (e.g., 3.0T) exploiting its higher signal to noise ratio. The BALI can be used to evaluate such brain structural changes in relation to their combined effect on cognition. Given how much existing data on AD and aging used conventional MRI (e.g., 1.5T), we examined the validity of using the BALI with low-field MRI data. Data were obtained from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Subjects who had MRI scans with T1 and T2-weighted images (T1WI and T2WI) at both 1.5T and 3.0T on the same day were retrieved (AD=37, MCI=45, HC=45). Clinical assessments included the Mini-Mental State Examination (MMSE) and the Alzheimer's disease Assessment Scale-cognitive subscale (ADAS-cog), completed within 14 days of neuroimaging. All images were evaluated applying the same BALI rating schema, to assess lesions in the deep white matter, periventricular, basal ganglia and the infratentorial regions, deficits in the cortical gray matter, the extent of dilated small vessels, and global atrophy. Maximum possible BALI=25; higher scores indicate greater damage. The inter-rater agreement rate was consistently high, ranging from 0.96 to 0.98. Under each field-strength and image-type condition, the BALI scores differed significantly between diagnosis (p<0.05): people in the AD group had the greatest BALI on average (12.7±3.0), whereas those in the healthy control group had the lowest (10.5±2.6). Under each condition, the BALI score was significantly correlated with age (r>0.35, p<0.001), MMSE (r>0.40, p<0.001), and ADAS-cog (r>0.36, p<0.001). The BALI score at 3.0T was slightly greater than that at 1.5T (F>3.90, p<0.06), so as with T2WI comparing to T1WI (F>3.99, p<0.05) without interaction (F<0.06, p>0.80). The BALI scores at 1.5T and 3.0T were significantly correlated, for both T1WI (r=0.94, p<0.001) and T2WI (r>0.95, p<0.001). T1WI and T2WI based BALI scores at 1.5T can be used to capture global brain structural changes and their relations to cognition.
What problem does this paper attempt to address?