Arduino, A; Pennecchi, F; Katscher, U; Cox, M; Zilberti, L (2023) Repeatability and Reproducibility Uncertainty Assessment in Magnetic Resonance-based Electric Properties Tomography of a Homogeneous Phantom. Tomography, 9 (1). pp. 420-435.

Preview

Text EID9691.pdf - Published Version Available under License Creative Commons Attribution. Download (1MB) | Preview

Abstract

In this paper, the uncertainty in the electric conductivity maps retrieved by magnetic resonance-based electric properties tomography (EPT), in particular the phase-based Helmholtz-EPT technique, applied on data acquired for a homogeneous phantom is assessed under repeatability and reproducibility conditions. Since phase-based Helmholtz-EPT elaborates linearly the scans acquired with magnetic resonance imaging, the law of propagation of uncertainty is used with no approximation to assess the uncertainty in the conductivity maps starting from the uncertainty in the magnetic resonance imaging scans. To quantify the uncertainty in the acquired scans, the robust James--Stein shrinkage estimator of the covariance matrix is used. Repeatable errors, not due to the phase-based approximation of the adopted Helmholtz-EPT technique, are recognized in the estimated conductivity maps and their dependence on the operational parameters of phase-based Helmholtz-EPT is quantified. Moreover, the spatial dispersion of the estimated electric conductivity maps is found to be very well described by the uncertainty evaluated under reproducibility conditions. The outcomes of the presented investigation suggest the possibility to compute the reproducibility uncertainty of a phase-based Helmholtz-EPT result from the spatial dispersion observed in homogeneous regions. In addition, it is shown that, in a homogeneous region, spatial averaging of the retrieved local conductivity values is quantitatively accurate. This paper provides a metrologically sound basis for the post-processing of the electric conductivity retrieved in homogeneous tissues adopted in some clinical studies, where the geometrical shape of the target tissue is identified by segmenting the magnetic resonance imaging scans.

Item Type:	Article
Keywords:	Magnetic resonance imaging, Uncertainty, Electromagnetic fields, Phantoms, Metrology
Subjects:	Mathematics and Scientific Computing > Modelling
Divisions:	Data Science
Identification number/DOI:	10.3390/tomography9010034
Last Modified:	02 May 2023 09:41
URI:	http://eprintspublications.npl.co.uk/id/eprint/9691

Actions (login required)

View Item