Perfusion-related parameters in intravoxel incoherent motion MR imaging compared with CBV and CBF measured by dynamic susceptibility-contrast MR technique

Citation
R. Wirestam et al., Perfusion-related parameters in intravoxel incoherent motion MR imaging compared with CBV and CBF measured by dynamic susceptibility-contrast MR technique, ACT RADIOL, 42(2), 2001, pp. 123-128
Citations number
32
Categorie Soggetti
Radiology ,Nuclear Medicine & Imaging
Journal title
ACTA RADIOLOGICA
ISSN journal
02841851 → ACNP
Volume
42
Issue
2
Year of publication
2001
Pages
123 - 128
Database
ISI
SICI code
0284-1851(200103)42:2<123:PPIIIM>2.0.ZU;2-C
Abstract
Objective. Perfusion-related parameters obtained by intravoxel incoherent m otion (IVIM) MR imaging (MRI) were compared with cerebral blood volume and how (CBV and CBF), retrieved by dynamic susceptibility-contrast (DSC) MRI. Material and Methods: Twenty-eight volunteers (average age 68.5 years) were investigated. Spin-echo echo-planar imaging with IVIM-encoding gradients w as employed (36 different b values, 0-1200 s/mm(2)). The perfusion fraction and the pseudo-diffusion coefficient were calculated for regions in thalam us gray matter and frontal white matter, using asymptotic and full fitting. In DSC-MRI, a Gd-DTPA-BMA contrast-agent bolus was monitored using simulta neous-dual FLASH. Deconvolution of the measured tissue concentration-versus -time curve with an arterial input function from the carotid artery was app lied, and maps of CBV and CBF were calculated. Results: The correlation between the perfusion fraction and CBV was r=0.56 (p<0.0000006) using asymptotic fitting, and r=0.35 (p<0.0004) when full fit ting was applied. Average CBF was 41.5 ml/(min 100 g), to be compared with the IVIM-based value of 63.6 ml/(min 100 g), obtained from the median value of the pseudo-diffusion coefficient in combination with assumptions about capillary network structure. Conclusion: The IVIM concept provided results that agreed reasonably with c onventional CBV and CBE The non-linear fitting to noisy signal data was pro blematic, in accordance with previously presented simulations.