Charles A. McKenzie, PhD Daniel Lim, MD Bernard J. Ransil, MD, PhD Martina Morrin, MD Ivan Pedrosa, MD Ernest N. Yeh, M Eng Daniel K. Sodickson, MD, PhD Neil M. Rofsky, MD Index terms: Magnetic resonance (MR), comparative studies, ** .121412, 2 ** .121416, ** .12143 Magnetic resonance (MR), reconstruction algorithms, ** .121412, ** .121416, ** .12143 Magnetic resonance (MR), technology, ** .121412, ** .121416, ** .12143 Magnetic resonance (MR), three- dimensional, ** .121412, ** .121416, ** .12143 Published online before print 10.1148/radiol.2302021230 Radiology 2004; 230:589 –594 Abbreviations: PARS = parallel imaging with an augmented radius in k space SENSE = sensitivity encoding SMASH = simultaneous acquisition of spatial harmonics VIBE = volumetric interpolated breath- hold examination 1 From the Departments of Radiology (C.A.M., D.L., M.M., I.P., N.M.R.), Medicine (Cardiovas- cular Division) (D.K.S.), and Neurology (B.J.R.), Beth Israel Deaconess Medical Center, 330 Brookline Ave, Room AN-239, Boston, MA 02215; Harvard Medical School, Boston, Mass (C.A.M., D.L., M.M., I.P., D.K.S., N.M.R.); and Harvard-MIT Division of Health Sciences and Technology, Boston, Mass (E.N.Y., D.K.S.). Received October 7, 2002; revision requested December 3; final revision received June 2, 2003; accepted June 25. D.K.S. supported by National In- stitutes of Health grants R29 HL60802 and R01 EB00447 and by a Whitaker Founda- tion Biomedical Engineering grant. Ad- dress correspondence to C.A.M. (e-mail: charles_mckenzie@caregroup.harvard.edu). 2 ** . Multiple body systems Author contributions: Guarantor of integrity of entire study, C.A.M.; study concepts, D.K.S., C.A.M.; study design, D.K.S., C.A.M., N.M.R.; liter- ature research, C.A.M.; clinical studies, C.A.M., D.L.; data acquisition, C.A.M., D.L.; data analysis/interpretation, C.A.M., D.L., I.P., M.M., E.N.Y; statistical analysis, B.J.R., C.A.M.; manuscript preparation and editing, C.A.M.; manuscript definition of intellectual content and final version ap- proval, C.A.M., D.K.S., N.M.R.; manuscript revision/review, all authors © RSNA, 2003 Shortening MR Image Acquisition Time for Volumetric Interpolated Breath-hold Examination with a Recently Developed Parallel Imaging Reconstruction Technique: Clinical Feasibility 1 A recently developed parallel mag- netic resonance (MR) imaging tech- nique, parallel imaging with an aug- mented radius in k space, was used to accelerate the volumetric interpo- lated breath-hold examination (VIBE) performed in 20 patients referred for clinical liver imaging. Nonaccelerated MR images were also acquired in these patients. A five-point scale was used to score the quality of the images. The acceleration resulted in reduced im- age quality: The nonaccelerated im- ages had a significantly higher (P  .05) mean score—3.8  0.3 (SD), indicating good quality—than the accelerated images—3.0  0.3, indi- cating acceptable quality. However, for three patients who could not hold their breath for the duration neces- sary for nonaccelerated imaging, less severe breathing artifacts on the ac- celerated images resulted in improved quality compared with the quality of the nonaccelerated images. Parallel MR imaging–accelerated VIBE may be beneficial for patients who have difficulty sustaining a breath hold for the duration necessary to perform nonaccelerated imaging. © RSNA, 2003 The volumetric interpolated breath-hold examination (VIBE) sequence (1,2) can yield fat-saturated three-dimensional mag- netic resonance (MR) images of the entire liver with high spatial resolution in all three dimensions. This technique has been shown to be reproducibly reliable for clinical evaluations of the upper part of the abdomen (2). Breath-hold times of longer than 20 seconds are often neces- sary for this technique, but some patients who are referred for liver imaging are not capable of holding their breath this long. The results of previous works have shown that parallel MR imaging tech- niques such as simultaneous acquisition of spatial harmonics (SMASH) (3) and sensitivity encoding (SENSE) (4) can be used to dramatically shorten the time needed for data acquisition while pre- serving the spatial resolution and con- trast of the original image. However, all parallel MR imaging techniques yield ac- celerated images with a lower signal-to- noise ratio—and possibly more artifacts— than the equivalent nonaccelerated images. Parallel MR imaging enables acceler- ated data acquisition by allowing one to skip the acquisition of the phase-encod- ing lines that would ordinarily be in- cluded in a nonaccelerated data acquisi- tion. If the accelerated images were to be reconstructed in the same manner as the nonaccelerated reference images, the re- duced phase encoding in the accelerated acquisitions would result in a reduced field of view and in images with substan- tial aliasing artifacts. A variety of parallel MR image reconstruction techniques to remove these aliasing artifacts from ac- celerated images have been developed. In general, this is accomplished by using knowledge of the spatial sensitivity pro- files of the radiofrequency coils used for 589 R adiology