Yonas Geda2, Mathew Senjem1, Prashanthi Vemuri1, Ronald Petersen1,. Clifford Jack1, 1Mayo Clinic, Rochester, Minnesota, United States; 2Mayo. Clinic ...
Alzheimer’s Imaging Consortium Posters: IC-P DTI study in Dementia (EDSD) we have collected data of more than 330 subjects from ten scanners located at nine sites. Objective: To assess the accuracy of ML classifiers for the detection of AD based on a large multicenter DTI data set using different approaches to reduce inter-site variability. Methods: After strict quality control we pooled the remaining 280 DTI and MRI scans derived from 137 patients with clinically probable AD and 143 healthy elderly controls. For classification we used fractional anisotropy (FA) maps and mean diffusivity (MD) maps and performed a tenfold cross validation. We selected discriminative voxels using the information gain criterion and classified the data with a Support Vector Machine. In a second step, we eliminated variance attributable to center and other covariates including age, education, gender, using principal component analysis (PCA) before repeating the classification procedure. Results: For FA and MD the feature selection identified areas in the medial temporal lobe and corpus callosum that had the strongest contribution to the group separation. We achieved an accuracy of 80% for FA and 83% for MD. For the tissue density maps we obtained 83% for WM and 89% for gray matter. The reduction of variance components arising from center, gender, age and education effects did not significantly change the classification results for FA and MD. Conclusions: Multicenter acquisition of DTI data in combination with multivariate ML approaches show promising results which can be compared to earlier monocenter DTI studies. Variance introduced by different scanners can be detected by PCA, but it seems not to affect the performance of the classifier.
IC-P-037
P25
that LOAD is associated with nucleus accumbens changes and EOAD with putamen abnormalities. These striatal abnormalities might reflect the distinct pattern of neuropathology in LOAD and EOAD patients (limbic vs neocortical atrophy). Caudate shape changes, which were found to be similar in EOAD and LOAD, might represent morphological reorganization due to ventricular enlargement.
MORPHOLOGICAL DIFFERENCES IN THE STRIATUM IN EARLY- AND LATE-ONSET ALZHEIMER’S DISEASE
Michela Pievani1, Martina Bocchetta1, Marina Boccardi1, Samantha Galluzzi1, Matteo Bonetti2, Paul Thompson3, Giovanni Frisoni4, 1 IRCCS Fatebenefratelli, Brescia, Italy; 2Istituto Clinico Citta di Brescia, Brescia, Italy; 3Laboratory of NeuroImaging, Los Angeles, California, United States; 4IRCCS Fatebenefratelli, Brescia, Italy.
Figure 1. In the putamen, both EOAD and LOAD showed a trend for reduced volumes (P 1.5. Area under the receiver operating characteristic curve (AUROC) evaluated the discrimination between PIB positive and negative subjects. Positive (PPV) and negative (NPV) predictive value was defined based on an estimated probability >0.50 who were PIB-positive. The estimated sample size for each characteristic, by age group (70-79 and 80-89 years), needed to screen to enroll 100 participants into a clinical trial with PIB>1.4 or >1.5 was determined based on the desired sample size divided by sample proportions in the MCSA data. Results: Of 483 CN individuals, 151 (31%) had PIB>1.5 and 211 (44%)>1.4. In univariate and multivariate models, discrimination was modest (AUROCw0.6-0.7). Multivariately, age and APOE best predicted odds of PIB>1.4 and >1.5. For PIB>1.5, the addition of all factors resulted in a PPV of 60% and NPV of 74%, and reduced the number needed to screen from 320 to 166 to enroll 100 individuals into a pre-clinical AD trial requiring brain amyloidosis. The predictability of some factors varied with age. For example, based on PIB>1.5, information on APOE genotype alone reduced the number needed to screen by 48% in persons aged 70-79 and 33% in those aged 80-89. Conclusions: Age and APOE genotype are useful predictors of amyloid accumulation, but discrimination is modest. Nonetheless, these results suggest that inexpensive and non-invasive measures could significantly reduce the number of CN individuals needed to screen with amyloid PET imaging or a lumbar puncture for CSF to identify a given number of amyloid positive subjects.
IC-P-039
REGIONAL CORTICAL THINNING PREDICTS WORSENING APATHY AND HALLUCINATIONS IN MILD COGNITIVE IMPAIRMENT AND MILD ALZHEIMER’S DISEASE
Nancy Donovan1, Lauren Wadsworth2, Natacha Lorius3, Joseph Locascio4, Dorene Rentz5, Keith Johnson6, Reisa Sperling7, Gad Marshall7, 1Center for Alzheimer Research and Treatment, Brigham and Women’s Hospital, Harvard Medical School and Cambridge Health Alliance, Boston, Massachusetts, United States; 2Massachusetts General Hospital, Charlestown, Massachusetts, United States; 3Center for Alzheimer Research and Treatment, Brigham and Women’s Hospital, Boston, Massachusetts, United States; 4Massachusetts General Hospital, Boston, Massachusetts, United States; 5Center for Alzheimer Research and Treatment, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, United States; 6Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, United States; 7Center for Alzheimer Research and Treatment, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts, United States. Background: Apathy and hallucinations are debilitating neuropsychiatric symptoms accompanying cognitive and functional decline in mild cognitive impairment (MCI) and Alzheimer’s disease (AD) dementia. Prior cross-sectional studies of apathy in AD dementia have most consistently implicated structural and functional changes in anterior cingulate and medial orbitofrontal cortices. The pathophysiological basis for hallucinations in AD is poorly understood. The objective of this study was to examine magnetic resonance imaging (MRI) cortical thickness and cerebrospinal fluid (CSF) AD biomarkers in relation to apathy and hallucinations, cross-sectionally and longitudinally, in a continuum of individuals with normal cognition (NC), MCI, and mild AD dementia. Methods: Eight hundred and twelve subjects from the Alzheimer’s Disease Neuroimaging Initiative study (229 NC, 395 MCI, 188 AD) underwent structural MRI at baseline and clinical assessments at baseline and longitudinally up to 3 years. CSF abeta, total tau, and phospho-tau were obtained for a subset of 413 subjects at baseline.
Backward elimination mixed random/fixed coefficient longitudinal regression models were used to evaluate the relationships between baseline cortical thickness in 6 regions (anterior cingulate, medial orbitofrontal, dorsolateral prefrontal, supramarginal, inferior temporal, occipital) and CSF biomarkers versus change in apathy and hallucinations measured by the Neuropsychiatric Inventory-Questionnaire. Covariates included the baseline dependent variable, diagnosis, gender, age, Apolipoprotein E, premorbid intelligence, memory performance, executive function, antidepressant use, and AD duration. General linear regression models were used to examine analogous cross-sectional associations at baseline. Results: Reduced baseline inferior temporal cortical thickness was predictive of increasing apathy over time (P