Quantitative diffusion imaging with steady-state free precession

• Published in: Magnetic resonance in medicine 2004-02, Vol.51 (2), p.428-433
• Main Authors: Deoni, Sean C.L., Peters, Terry M., Rutt, Brian K.
• Format: Article
• Language: English
• Subjects: Algorithms; apparent diffusion coefficient; Brain Mapping - methods; diffusion imaging; Diffusion Magnetic Resonance Imaging - methods; Image Processing, Computer-Assisted; Models, Statistical; Phantoms, Imaging; rapid volumetric imaging; steady-state free precession
• ISSN: 0740-3194; 1522-2594
• DOI: 10.1002/mrm.10708

The addition of a single, unbalanced diffusion gradient to the steady‐state free precession (SSFP) imaging sequence sensitizes the resulting signal to free diffusion. Unfortunately, the confounding influence of both longitudinal (T1) and transverse (T2) relaxation on the diffusion‐weighted SSFP (dwSSFP) signal has made it difficult to quantitatively determine the apparent diffusion coefficient (ADC). Here, a multistep method in which the T1, T2, and spin density (Mo) constants are first determined using a rapid mapping technique described previously is presented. Quantitative ADC can then be determined through a novel inversion of the appropriate signal model. The accuracy and precision of our proposed method (which we term DESPOD) was determined by comparing resulting ADC values from phantoms to those calculated from traditional diffusion‐weighted echo planar imaging (dwEPI) images. Error within the DESPOD‐derived ADC maps was found to be less than 3%, with good precision over a biologically relevant range of ADC values. Magn Reson Med 51:428–433, 2004. © 2004 Wiley‐Liss, Inc.