Meeting Banner
Abstract #2247

Quality Assessment and Consistency of Serial Multi-Center Resting-State fMRI in Alzheimers Disease

David Clayton1, Farshid Faraji1, Douglas Arnold2, Alex De Crespigny3, Alexandre Coimbra3

1Genentech, Inc., South San Francisco, CA, United States; 2NeuroRX Research, Montreal, Quebec, Canada; 3Genentech Inc, South San Francisco, CA, United States

Several single site resting-state fMRI studies (RS-fMRI) have shown that brain functional connectivity metrics are altered in Alzheimer Disease (AD) patients. Exploratory RS-fMRI was included in two global Phase II studies of a novel AD drug. We report here fMRI data quality control procedures and observed quality metrics from screening and very early on-treatment scans. Subject tolerance and protocol compliance was good. In spite of heterogeneous scanning equipment and variable raw SNR across sites, the data pre-processing pipeline resulted in consistent processed data quality metrics, with few scans that could not be used for subsequent functional connectivity analysis.

Keywords

accompanied acquisition allowed altered analyzed applied asked asleep assessed assessment assumption automated available background brain centrally clinical closed coefficient coils compliance connectivity consecutive consisted consistency consistent consistently control controlled corners corrected correction correlation coverage cutoff datasets decreasing defined described development deviation deviations discouraged disease displacement drug equivalent exception experience exploratory extracted eyes facilitate filtered filtering function functional functions general global good gradient gray head heterogeneous highly house implemented included intensity inter interval intervals intra kernel linear locally made magnets majority manufacturers many mask materials matrix median metrics minimal minor moderate motion noise normalized novel nuisance occurred pairs patients pipeline predefined processing progression protocol quality received registered registration regressed regressing relatively reported required resolution rest resting retest saturation screening selected serial sets seven several shimming shot simple site sites slice slices south spatial spite spoiled step studies subject subsequent superior tell temporal temporally therapy tool treatment trial valuable variable variables variation whether