Skip to main content
 

Comparison of QSM estimation methods using multiple head orientations

Friday, September 16, 2016 — Poster Session IV

12:00 p.m. – 1:30 p.m.
FAES Terrace
CC
BIOENG-2

Authors

  • W-T Wang
  • H Ravi
  • D Pham
  • JA Butman

Abstract

Quantitative susceptibility mapping (QSM) is a recently developed MRI technique for quantifying the spatial distribution of magnetic susceptibility within biological tissues. Magnetic susceptibility differs among tissues based on their molecular composition and is a direct reflection of cellular architecture of the tissue. Due to its sensitivity to molecular composition, QSM has been used to investigate quantification of tissue iron concentration, several pathologies such as cerebral micro-bleeds, Parkinson's diseases, and multiple sclerosis and blood oxygen saturation. Susceptibility modifies the magnetic field detected in the MR signal phase. The determination of an arbitrary susceptibility distribution from the induced field shifts is an inverse problem that is challenging due to its ill-posedness. A multi-orientation method, calculation of susceptibility through multi-orientation sampling (COSMOS) is proposed to stabilize this inverse problem. The field created by the susceptibility distribution is sampled at multiple orientations with respect to the polarization field B0, and the susceptibility map is reconstructed. However, multiple orientation data would be time consuming to acquire which may lead to patient discomfort and may not be feasible for routine clinical use. Conversely, susceptibility maps from a single acquisition with the head in one position could be calculated using thresholded k-space division (TKD). However, this technique suffers from poor quality compared to COSMOS. In this work, we compared the performance of COSMOS QSM with TKD QSM averaged at variable number of orientations. We found that for fewer number of orientations (< 4 orientations) averaged TKD QSM performs better than COSMOS.

Category: Biomedical Engineering and Biophysics