MRI analysis of medial temporal lobe atrophy vs. psychometric biomarkers of Alzheimer's disease


Psychometric biomarkers have been used to predict Alzheimer’s disease (AD), but due to subjectivity and ambiguity, they can lack accuracy. Neuroimaging such as magnetic resonance imaging (MRI) provides a fast, objective, and effective means to monitor AD progression. The development of AD occurs in stages; however, the most salient clinical manifestation is a decline in cognitive functioning. To indicate which early screening technique more accurately predicts cognitive decline and further progression to AD, the specificity and sensitivity of psychometric versus MRI biomarkers is compared. In this study, we show that specific tissue loss assessed by anatomical scans in the medial temporal lobe (MTL) occurs early in AD pathology and that analysis of this atrophy can be used to predict future clinical and cognitive decline in normal subjects (NL) and subjects with mild cognitive impairment (MCI). Stable subjects (NL–>NL: n=32) and decliners (NL–>MCI: n=22, NL–>AD: n=4, and MCI–>AD: n=2) completed a standardized longitudinal protocol including history, physical, clinical laboratory with ApoE4 allele screening, neurological and neuropsychological evaluations, and two high-resolution MRI examinations. Observation period ranged from 5.1 to 13.2 years. Baseline and follow-up MRI images were co-registered and the annual rate of brain tissue atrophy within the MTL and whole brain was calculated using the boundary shift algorithm. Binary logistic regression was applied to calculate sensitivity, specificity, and accuracy of neuroimaging and psychometric predictors. The results corroborate the application of MRI analysis techniques as surrogate markers of Alzheimer’s disease; however, validation of this technique, with postmortem AD case verification, is warranted.