Abdullah S Bdaiwi1,2, Peter J Niedbalski1, Md M Hossain3, Matthew M Willmering1, Laura L Walkup1,2,4, Hui Wang5, Robert P Thomen1, Kai Ruppert1, Jason C Woods1,4, Zackary I Cleveland1,2,4. 1. Center for Pulmonary Imaging Research, Division of Pulmonary Medicine, Children's Hospital Medical Center, Cincinnati, Ohio, USA. 2. Department of Biomedical Engineering, University of Cincinnati, Cincinnati, Ohio, USA. 3. Division of Biostatistics and Epidemiology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio, USA. 4. Department of Pediatrics, University of Cincinnati, Cincinnati, Ohio, USA. 5. Philips Healthcare, Cincinnati, Ohio, USA.
Abstract
RATIONALE: Hyperpolarized (HP) 129 Xe-MRI provides non-invasive methods to quantify lung function and structure, with the 129 Xe apparent diffusion coefficient (ADC) being a well validated measure of alveolar airspace size. However, the experimental factors that impact the precision and accuracy of HP 129 Xe ADC measurements have not been rigorously investigated. Here, we introduce an analytical model to predict the experimental uncertainty of 129 Xe ADC estimates. Additionally, we report ADC dependence on age in healthy pediatric volunteers. METHODS: An analytical expression for ADC uncertainty was derived from the Stejskal-Tanner equation and simplified Bloch equations appropriate for HP media. Parameters in the model were maximum b-value (bmax ), number of b-values (Nb ), number of phase encoding lines (Nph ), flip angle and the ADC itself. This model was validated by simulations and phantom experiments, and five fitting methods for calculating ADC were investigated. To examine the lower range for 129 Xe ADC, 32 healthy subjects (age 6-40 years) underwent diffusion-weighted 129 Xe MRI. RESULTS: The analytical model provides a lower bound on ADC uncertainty and predicts that decreased signal-to-noise ratio yields increases in relative uncertainty ( ϵ ADC ) . As such, experimental parameters that impact non-equilibrium 129 Xe magnetization necessarily impact the resulting ϵ ADC . The values of diffusion encoding parameters (Nb and bmax ) that minimize ϵ ADC strongly depend on the underlying ADC value, resulting in a global minimum for ϵ ADC . Bayesian fitting outperformed other methods (error < 5%) for estimating ADC. The whole-lung mean 129 Xe ADC of healthy subjects increased with age at a rate of 1.75 × 10-4 cm2 /s/yr (p = 0.001). CONCLUSIONS: HP 129 Xe diffusion MRI can be improved by minimizing the uncertainty of ADC measurements via uncertainty propagation. Doing so will improve experimental accuracy when measuring lung microstructure in vivo and should allow improved monitoring of regional disease progression and assessment of therapy response in a range of lung diseases.
RATIONALE: Hyperpolarized (HP) 129 Xe-MRI provides non-invasive methods to quantify lung function and structure, with the 129 Xe apparent diffusion coefficient (ADC) being a well validated measure of alveolar airspace size. However, the experimental factors that impact the precision and accuracy of HP 129 Xe ADC measurements have not been rigorously investigated. Here, we introduce an analytical model to predict the experimental uncertainty of 129 Xe ADC estimates. Additionally, we report ADC dependence on age in healthy pediatric volunteers. METHODS: An analytical expression for ADC uncertainty was derived from the Stejskal-Tanner equation and simplified Bloch equations appropriate for HP media. Parameters in the model were maximum b-value (bmax ), number of b-values (Nb ), number of phase encoding lines (Nph ), flip angle and the ADC itself. This model was validated by simulations and phantom experiments, and five fitting methods for calculating ADC were investigated. To examine the lower range for 129 Xe ADC, 32 healthy subjects (age 6-40 years) underwent diffusion-weighted 129 Xe MRI. RESULTS: The analytical model provides a lower bound on ADC uncertainty and predicts that decreased signal-to-noise ratio yields increases in relative uncertainty ( ϵ ADC ) . As such, experimental parameters that impact non-equilibrium 129 Xe magnetization necessarily impact the resulting ϵ ADC . The values of diffusion encoding parameters (Nb and bmax ) that minimize ϵ ADC strongly depend on the underlying ADC value, resulting in a global minimum for ϵ ADC . Bayesian fitting outperformed other methods (error < 5%) for estimating ADC. The whole-lung mean 129 Xe ADC of healthy subjects increased with age at a rate of 1.75 × 10-4 cm2 /s/yr (p = 0.001). CONCLUSIONS: HP 129 Xe diffusion MRI can be improved by minimizing the uncertainty of ADC measurements via uncertainty propagation. Doing so will improve experimental accuracy when measuring lung microstructure in vivo and should allow improved monitoring of regional disease progression and assessment of therapy response in a range of lung diseases.
Authors: Harald E Möller; X Josette Chen; Brian Saam; Klaus D Hagspiel; G Allan Johnson; Talissa A Altes; Eduard E de Lange; Hans-Ulrich Kauczor Journal: Magn Reson Med Date: 2002-06 Impact factor: 4.668
Authors: Rafael L O'Halloran; James H Holmes; Talissa A Altes; Michael Salerno; Sean B Fain Journal: J Magn Reson Date: 2006-12-05 Impact factor: 2.229
Authors: Samuel Patz; F William Hersman; Iga Muradian; Mirko I Hrovat; Iulian C Ruset; Stephen Ketel; Francine Jacobson; George P Topulos; Hiroto Hatabu; James P Butler Journal: Eur J Radiol Date: 2007-09-24 Impact factor: 3.528
Authors: D Guenther; B Eberle; J Hast; J Lill; K Markstaller; M Puderbach; W G Schreiber; G Hanisch; C P Heussel; R Surkau; T Grossmann; N Weiler; M Thelen; H U Kauczor Journal: NMR Biomed Date: 2000-06 Impact factor: 4.044