Meeting Banner
Abstract #2292

Enhanced Phase Regression with Savitzky-Golay Filtering for High-Resolution BOLD FMRI

Robert L. Barry1, 2, John C. Gore, 23

1Vanderbilt University Institute of Imaging Science, Nashville, TN, United States; 2Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, Nashville, TN, United States; 3Vanderbilt University, Nashville, TN, United States

Phase regression exploits the temporal evolution of phase in individual voxels to suppress BOLD fluctuations from larger vessels while preserving signal changes from microvascular effects. However, the efficacy of this algorithm is hindered when the phase time series exhibits low signal-to-noise ratio. We demonstrate that Savitzky-Golay filters may be used to recover the underlying change in phase and completely restore the efficacy of phase regression. This approach is shown to work on data acquired with single-shot and multi-shot pulse sequences, and should be useful for both human and animal gradient-echo fMRI at high spatial resolutions at high- and ultra-high fields.

Abstract PDF Poster PDF

Keywords

acquisition activation aggressive although analyses analytical applied approved authors averaging away blue board bold brain brief central channel clear clearly coil combinations complete considered containing contrast curve dependent displays draining efficacy eight either eleven enhance enhanced every evolution example exhibits fast field filtered filtering filters finally flashing form frame function functional gradient gray highest hindered histograms human hypothesis hypothesized identify imaginary impetus implementation increasing institutional john known larger largest linear magnitude majority many maps medical nearly negligible neural noise noisy observe optimal oxygenation paradigm partial physiological planned polynomial positive predominantly preserves preserving presto previously primary protocol pulse quadrature radiological radiology readily reduce reflect regarding registered regression resolution resolutions review runs sans scanner science sciences scores segments series several shot significantly smoothing spatial specificity specified stability steady still strategy strength subject subjects suppress suppressed suppresses suppression temporal thereby thick ultra underlying undesirable unfiltered upon validate variance vary veins vessel vessels volume volunteers