Abstract |
|
Until now, in the Magnetic resonance imaging image quality assessment using JPEG2000 software,
phantom images have been mostly assessed by the compression rate, not considering SENSE factors.
In this study, performed on a Philips 3.0-Tesla scanner, we took images for various SENSE factors
(1, 1.5, 2, 2.5, and 3) and used the ACR MRI phantom system, largely used worldwide. We
performed lossy compression of the individual images by using the JPEG2000 software to evaluate
those images (Q-factor = 5:1, 10:1, 20:1, and 40:1, respectively). According to the ACR MRI QC
manual, the geometric accuracy of the phantom images did not depend on the SENSE factors and
on the lossy image compression rate. Without any change induced by the SENSE factors and on
the lossy image compression rate, the high contrast spatial resolution measured 1.0 mm in T1 TSE.
The slice thickness accuracy was measured at 4.36 mm, 4.36 mm, and 4.32 mm in the original,
5:1 and 10:1 compressions of the T1 TSE, respectively, without any change in the SENSE factors.
However, in the 20:1 lossy compression, it was reported at 4.28 mm, thus surpassing the baseline.
In the compressions of the T2 TSE, it measured on average 4.54 mm, 4.54 mm, 4.53 mm, and
4.46 mm in the original, 5:1, 10:1, and 20:1 compressions, respectively, without any change in the
SENSE factors. The slice position accuracy in slice 1 and slice 11 of the T1 and of the T2 TSE
measured 2.93 mm and 0.98 mm, respectively, without any dependence on the SENSE factors on
the lossy image compression rate. The present study suggests that a SENSE factor of more than
2.5 not be applied in T1 and T2 TSE and that lossy compression not be done where a SENSE of
1.5¡ more factor is given in the use of JPEG2000. |