Meeting Banner
Abstract #2715

Concurrent Magnetic Field Monitoring of EPI Time Series: Characterizing Encoding Field and Image Fluctuations Using Principal Component Analysis

Saskia Klein1, Lars Kasper1, Johanna Vannesjo1, Maximilian Hberlin1, Simon Gross1, Benjamin Dietrich1, Klaas P. Prssmann1

1Institute for Biomedical Engineering, ETH Zurich, Zurich, Switzerland

In fMRI, changes in image intensity are induced by signal fluctuations, e.g. related to the brain metabolism (BOLD). They can also be caused by gradient system instabilities and drifts, leading to confounds in time series analyses.

Keywords

according activation amount analyses analyzed applications apply artifacts asymmetric attributable attributed axial beneficial biomedical bold brain calibrate calibrated calibration caused certain channel characteristic characterized characterizing coefficient coefficients coil comparative component components computed concurrent considerable contained contribution correcting correlation corresponded correspondence correspondences correspondingly corresponds course days depends detect determine deviation dimensions disentangle domain doped downward drift drifts duration dynamics encoding engineering evolution explain field fitted fluctuation fluctuations former gain ghost gradient gross harmonics head highly history ignoring incomplete individually induced infer influences instabilities intensity iterative lasted latter lead linear mainly major materials measure measured minutes model models modulation modulations monitoring necessarily neglecting noise object patterns phantom pixel principal probes projection projections proportion protocol quantify quantitative random readout reconstructed reconstructing reconstruction reflect related represents resolution schemes sensitivity separated series session sets setup several slices space spatial sphere stimulus strong sufficient system systematic table trajectory typical variability variance vectors water whereas whose widening yielded