Premium
Reproducibility of the quantification of arterial and tissue contributions in multiple postlabeling delay arterial spin labeling
Author(s) -
Sousa Inês,
Vilela Pedro,
Figueiredo Patrícia
Publication year - 2014
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.24493
Subject(s) - reproducibility , cerebral blood flow , bayesian probability , biomedical engineering , medicine , nuclear medicine , biological system , computer science , statistics , mathematics , biology
Purpose To evaluate the reproducibility of estimation of cerebral blood flow (CBF), bolus arrival time (BAT), and arterial blood volume (aBV) from arterial spin labeling (ASL) data acquired at multiple postlabeling delays (PLDs). Materials and Methods CBF, BAT, and aBV parameters were estimated from flow‐suppressed and nonflow‐suppressed multiple‐PDL PICORE‐Q2TIPS ASL using model‐based Bayesian and least‐squares fitting frameworks, and aBV was also obtained from a model‐free approach. Reproducibility of these parameters was assessed by computing the within‐ and between‐subject coefficients of variability (CVw and CVb). Results CVw and CVb were comparable across model‐based approaches, but were greater for the aBV from the model‐free approach. Overall, the Bayesian model estimation procedure was found to provide the best compromise between reliability and reproducibility, yielding CVw/CVb values of 21/21, 3/4, and 24/26% for CBF, BAT, and aBV, respectively. Although a CBF range of 45 mL/100g/min to 59 mL/100g/min was found on average and a BAT of 0.7–1.0 seconds across methods, the corresponding maps were comparable in terms of the parameters' spatial distributions, and in particular in the identification of macrovascular locations, as assessed through comparison with time‐of‐flight images. Conclusion Reproducible estimates of CBF, BAT, and aBV values can be obtained from non‐macroflow‐suppressed ASL using both least‐squares and Bayesian model‐based methods. J. Magn. Reson. Imaging 2014;40:1453–1462 . © 2013 Wiley Periodicals, Inc .