The University of Sussex

The use of time series analysis for the study of respiratory motion artefacts in MRI/MRS: review and initial results from the use of the Lomb-Scargle periodogram

Theodoros N. Arvanitis, Des Watson

Various sources of systematic error exist for both Magnetic Resonance Imaging and Spectroscopy. In particular, MRI and MRS are sensitive to any type of microscopic and macroscopic motion. In this paper we study the use of the Lomb-Scargle periodogram in the analysis of k-space data from motion artefacted MR images and spectra, while we provide a detailed review on the theoretical foundations of this method. We investigate the usefulness of this technique for detecting suspected periodic and quasi-periodic perturbations in MR signals, introduced by respiratory motion. Our work strongly suggests that the Lomb-Scargle periodogram can offer interesting qualitative information, relating to the effect of motion artefacts in MRI/MRS. The paper discusses the issues relating to the future use of this method for the identification and analysis of systematic errors in clinical MRI/MRS.

Download compressed postscript file