Image UniformityHow do you measure image uniformity?
|
|
Scan of a homogeneous phantom, windowed to bring out intensity differences
|
Measurement of image uniformity is performed using a homogenous oil or ion-doped water phantom. The American College of Radiology (ACR) method is the simplest, in which the image is first windowed to bring out subtle intensity differences. The maximum (max) and minimum (min) intensity values are then measured for small (~ 1 cm²), and a Percent Image Uniformity (PIU) is calculated as
PIU = {1 − (max−min) / (max+min) } x 100%
where 100% indicates perfect uniformity.
|
I have always been surprised to find that even with the use of image intensity correction filters (such as PURE, SCIC, CLEAR, etc) typical PIU values on a properly calibrated modern 3.0T scanner may only be slightly above 85%. This means that when scanning a homogeneous phantom up to a 15% variation in pixel values across the image may normally be present. (This is in stark contrast to CT scanning, where very little variation in region-of-interest (ROI) values across a uniform phantom is seen). So be careful when you measure ROI differences in an MR image in a clinical setting; you may just be measuring non-uniformity!
Image uniformity problems can result from: 1) lack of centering -- i.e. the patient or phantom is too close to one side of the coil; 2) ghosting from motion or eddy currents; or 3) failure of coil components including decoupling circuitry.
References
American Association of Physicists in Medicine (AAPM). Report No.100. Acceptance testing and quality assurance procedures for magnetic resonance imaging facilities. 2010.
American College of Radiology. Phantom test guidance for use of the large MRI phantom for the ACR accreditation program, 2018.
National Electrical Manufacturers Association. NEMA Standards Publication MS 3-2008 (R2014). Determination of image uniformity in diagnostic magnetic resonance images. NEMA, Washington, DC. Can be downloaded for free or purchased in book form here.
American Association of Physicists in Medicine (AAPM). Report No.100. Acceptance testing and quality assurance procedures for magnetic resonance imaging facilities. 2010.
American College of Radiology. Phantom test guidance for use of the large MRI phantom for the ACR accreditation program, 2018.
National Electrical Manufacturers Association. NEMA Standards Publication MS 3-2008 (R2014). Determination of image uniformity in diagnostic magnetic resonance images. NEMA, Washington, DC. Can be downloaded for free or purchased in book form here.