Ryan Bosca MS, Presenter: Nothing to Disclose

Edward Ashton, Abstract Co-Author: Officer, VirtualScopics, Inc

Gudrun Zahlmann PhD, Abstract Co-Author: Employee, F. Hoffmann-La Roche Ltd

Edward F. Jackson PhD, Abstract Co-Author: Nothing to Disclose

The RSNA Quantitative Imaging Alliance (QIBA) DCE-MRI Technical Committee developed a DCE-MRI Phantom that can be used in site qualification, quality control, assessment of RF coil response characteristics, and data harmonization across vendors, centers, and time. The phantom design details, associated data analysis software, and initial results from test/retest T1 measurements and DCE-MRI acquisitions are presented.

The phantom consists of a water-filled cylindrical volume containing 2 sets of 3-cm diameter spheres. The first set of 8 spheres (distributed around a 14-cm circle) has T1 values ranging from 21 to 1333ms, representative of values measured in a vascular compartment before/following contrast agent injection. A second set of 24 spheres is distributed around a 29-cm circle. The 24 spheres represent 3 sets of 8, with each set rotated by approximately 90o and having T1 values ranging from 133-1500ms, representative of values measured in tissue before/following contrast agent injection. Data were acquired on 1.5T GE HDxt and 3.0T GE MR750 scanners. Images were acquired at baseline and 1 week later using an inversion recovery (IR) sequence (10 inversion times) and a multiple flip angle (N=7) fast spoiled gradient echo (FSPGR) sequence. FSPGR images were acquired twice, once using a phased array receive coil and once using the transmit/receive body coil. Data were analyzed using in-house software as well as freely available software developed specifically for use with the phantom. Bland-Altman repeatability and limits were computed.

FSPGR repeatability values ranged from 23-68ms and 15-62ms using the phased array and body coils, respectively, with limits of agreement ranging from -99 to 65ms and -71 to 67ms. IR repeatability values ranged from 2-13ms with limits of agreement ranging from -15 to 13ms. Coefficients of variation in T1 from the three pseudo-rotations of the ‘tissue’ spheres were in the range 12-39% (FSPGR) and 0-1% (IR).

The QIBA DCE-MRI phantom allows assessment of contrast response, T1 relaxometry reproducibility, and signal stability from a volume of interest relevant to body applications.

The RSNA QIBA DCE-MRI phantom, and associated analysis software, has been designed for the qualification and quality control of sites involved in clinical trial /clinical DCE-MRI quantitative imaging.

Bosca, R, Ashton, E, Zahlmann, G, Jackson, E, RSNA Quantitative Imaging Biomarker Alliance (QIBA) DCE-MRI Phantom: Goal, Design, and Initial Results.  Radiological Society of North America 2012 Scientific Assembly and Annual Meeting, November 25 - November 30, 2012 ,Chicago IL. http://archive.rsna.org/2012/12033723.html