The intent of the MRI Accreditation Program is to use the information obtained from the review of both clinical and phantom images to assess overall image quality. Your facility will need to perform two specified phantom scans using ACR protocols as well as a second set of phantom scans using your site’s routine clinical head protocol as outlined in these instructions.
Now that both medium and large phantoms are approved for accreditation of MR scanners in the modular Magnetic Resonance Accreditation Program (MRAP), sites must submit phantom images acquired using a head coil that is routinely used for clinical brain imaging on the scanner and must use the largest phantom that fits inside that head coil. Facilities with scanners that do not have a head coil and/or do not routinely perform brain imaging should use the small phantom in the knee coil to obtain phantom images for accreditation review.
The ACR Large and Medium MRI phantoms are hollow cylinders of acrylic plastic closed at both ends. Both phantoms are filled with a 10 mM NiCl2, 75 mM NaCl solution. The internal (signal producing) phantom dimensions are displayed in Table 1. Both phantoms contain resolution test objects consisting of either three-hole or four-hole patterns. Future large phantoms will include four resolution patterns. The outside of each phantom has the words “NOSE” and “CHIN” etched into it as an aid to orienting the phantom for scanning, as if it were a head.
Table 1: Large and Medium Phantom resolution patterns and internal (signal producing) dimensions.
Resolution Pattern (mm)
Internal Length (mm)
Internal Diameter (mm)
Head coils large enough to fit the large phantom
1.1, 1.0, 0.9
Smaller phased-array head coils
1.1, 1.0, 0.9, 0.8
Inside the phantoms are several structures designed to facilitate a variety of tests of scanner performance. A description of each structure is included in the relevant test section. The methods for making the measurements are generally the same for both phantoms. However, the limits and ROI sizes differ for certain tests.
Phantom Set-Up and Alignment for Scanning
The MRI Accreditation Large Phantom/Medium Phantom should be scanned in the head coil with the cylindrical phantom aligned as a head would be in the coil. Transaxial slices should result in circular cross-sections of the phantom. The phantom should be positioned so that the word “Nose” is where the nose would be for a standard head study and the word “Chin” is where the chin would be located in a standard head study. The center of the phantom (the dark notch on the side of the phantom) should be placed in the center of the head coil and aligned with the positioning indicator light so that it will be in the isocenter of the scanner. Once grossly positioned, it is then necessary to “fine tune” the position of the phantom along all three axes. For this, you will need to use a non-metallic, MRI safe bubble level, such as the one depicted below. Place the level along the top of the phantom running in and out of the scanner (along the z-axis) to ensure that the phantom is horizontal. A custom-made phantom holder that ensures the phantom is centered in the head coil or a stack of printer paper, adjusted to the proper height, may be required to ensure the phantom is positioned and levelled vertically.
Once correctly aligned, the phantom should be kept in the same position during the entire series of scans.
You may use a small, nonmagnetic bubble level (not provided) for positioning the phantom.
Figure 1: Example of small, nonmagnetic bubble level
Place the level on top of the plastic bar at the chin surface, rotating the phantom so that the plastic bar is horizontal. With the phantom then clamped or wedged inside the head coil, check to see that the sagittal laser alignment light is parallel to the line running along the “nose” surface of the phantom (to see the laser light reflection, it may be necessary to place a piece of white paper on top of the phantom). After each position adjustment, recheck that the top of the phantom and the chin bar are still horizontal. After the phantom has been moved into the center of the magnet, verify its positioning by performing sagittal and, if desired, coronal plane localizer scans, until correct.
It is not acceptable to cool the phantom before scanning, to improve SNR.
Phased array head coils naturally produce images that are less uniform due to the smaller coil elements, as compared to quadrature coils. Be sure to apply the vendor’s intensity correction to the ACR T1 and T2 series if they were acquired using a multi-channel phased array coil. The correction goes by different names depending on vendor (SCIC, PURE, CLEAR, Normalize, Pre-scan normalize, and B1 Filter are some examples).
It is not acceptable to use deep learning or artificial intelligence (DL or AI) reconstruction options for phantom submissions.
The Required Images
For both the large and medium phantoms, the phantom portion of the MRAP requires the acquisition of a sagittal localizer and four axial series of images. The same set of 11 slice locations within the phantom is acquired in each of the four, axial series of images. These images are acquired using the scanner’s routine clinical head coil with the largest ACR phantom that fits inside that coil. The scan parameters and approximate scan times for the sagittal localizer and the first 2 axial series of images are in Table 2. These three series are referred to as the ACR sequences or series. The third and fourth series of axial images are based on the site’s own clinical brain protocols and are referred to as the site sequences or site series.
Note: MRI systems require that a weight be entered in order to scan the phantom; the ACR
recommends that your site enter a weight of 200 pounds. For some 3T scanners, using a weight
of 200 lbs. will cause the ACR T1 series scan time to double from 2:16 to over 4 minutes.
The scan time doubles to meet head SAR model restrictions. You could reduce the
entered weight to approximately 50 lbs. to maintain the 2:16 acquisition time. The image quality
should not be affected. Note that this is only a workaround for phantom scans. For patient scans the actual patient weight should always be used to ensure patient safety.
The localizer is a sagittal spin-echo acquisition through the center of the phantom, that is referred to as the sagittal localizer. In mid-2021 the thickness was reduced to 10mm. However, sites may continue to use 20mm slice thickness if they prefer.
The first axial series is a spin-echo acquisition with ACR-specified scan parameters that are typical of T1- weighted acquisitions. This series is called the ACR T1W series.
Prior to mid-2021, the second axial series was a double spin-echo acquisition with ACR-specified scan parameters that were typical of proton density/T2-weighted acquisitions common at the time the MRAP began. Since the double-echo sequence is rarely used in modern imaging, starting in mid-2021 the ACR T2W series has been changed to a single echo spin echo with the same TR and TE as the double echo T2. For the ACR T2W series, sites are encouraged to switch to a single echo spin echo, but still have the option of submitting the double echo series. When analyzing data from a double-echo acquisition, only the second-echo images (TE=80) are evaluated.
The third and fourth axial series are based on the scan parameters the site normally uses in its clinical protocols for axial T1- and T2-weighted brain studies, respectively. These series are called the site T1W and site T2W series. Sites should scan the phantom using their clinical protocols, but may change the number of slices, slice thickness, and interslice gap to the values in Table 2.
It has come to our attention that some manufacturers of MRI systems have sent “sample” or “recommended” phantom site scanning protocols to their users. Please be aware that the requirement for MRI accreditation is that for the site T1 and site T2 series, facilities must use the same protocol (with appropriate modifications) for the phantom that the facility uses for routine T1 and T2 brain imaging. Failure to comply with this requirement could result in failure to achieve accreditation.
Table 2: Large and Medium Phantom scan parameters and approximate scan times. If scan times are significantly different, check scan parameters to ensure they are correct.
FOV (mm) (frequency)
FOV (mm) (phase)
Slice thickness (mm)
Slice gap (mm)
Scan Time (min:sec)
*ACR Sag localizer
ACR Axial T1
ACR Axial T2
***Site Axial T1 Brain
***Site Axial T2 Brain
*For the ACR Sag localizer 10mm slice thickness is preferred, but 20mm is acceptable.
**For the ACR T2 series single echo spin echo is preferred, but double echo spin echo (TR 2000, TE 20/80) is acceptable. Fast/Turbo spin echo must not be used.
***Blank fields indicate where to use the site’s clinical parameters from routine brain protocols.
†For 3DFT clinical sequences only
Each axial series has 11 required slice locations. The locations are numbered starting at the inferior end of the phantom; so, slice location 1 is at the end of the phantom labeled “CHIN.” The phantom should be scanned inferior to superior. However, even if the images are acquired in reverse order, this document will refer to them by their series name and slice location number. For example, ACR T1 slice 7 is the image at slice location 7 of the ACR Axial T1-weighted acquisition.
For all four axial series image sets, the required slice thickness is 5 mm and the slice gap is 5 mm. Thus, the set of 11 slices spans a distance of 100 mm from the center of the first slice to the center of the last slice.
Figure 2 shows sagittal localizers of the large and medium phantoms with the 11 axial slice locations cross- referenced. There are two pairs of crossed 45° wedges lying in the central sagittal plane of the phantom: one pair at each end of the phantom. Slice 1 is prescribed to be centered on the vertex of the angle formed by the crossed wedges at the inferior end of the phantom. The vertices of the wedge pairs are separated by 100 mm, and therefore slice 11 falls on the vertex at the superior end of the phantom.
When replicating 3DFT site protocols in the phantom, the operator should prescribe 21 slices of 5mm thickness to ensure that Slice 1 is centered on the vertex of the angle formed by the crossed wedges at the inferior end of the phantom and slice 21 falls on the vertex at the superior end of the phantom. In this case the images to be analyzed will be images 1, 3, 5, 7, 9, 11,13, 15, 17, 19, and 21. The others are just “gap” slices.
The phantom image data must be electronically uploaded to the ACR in uncompressed or lossless compressed DICOM format. Do not upload images in lossy compressed format or the image data may be rejected. Detailed submission information can be found here: Testing Package and Image Submission: Overview.
Figure 2: Sagittal localizers of the Large (left) and Medium (right) phantoms showing the 11 required axial slice locations and the paired 450 wedges. The words “CHIN” and “NOSE” indicate where those words are etched into the phantoms as an aid to orienting them for scanning as if they were a head.
Available vendor phantom order forms include:
JM Specialties Large Phantom order form: See attached below. (Note, due to a current backlog, JM Specialties is not accepting orders for the ACR Medium Phantom at this time. For the medium phantom, please access the link below to the Diagnomatic online store.
Revision History for this Article
Description of Revision(s)
Article created; FAQs incorporated; No criteria changes
Evaluating Phantom Image Quality
Guidance and Order forms attached
Added Medium phantom
Added Large phantom order form attachment
Updated Large and Medium phantom order form attachments
Phantom Set-Up and Alignment for Scanning
Added DL/AI reconstruction language
Phantom Order Forms
Added link to Diagnomatic Medium Phantom Order Form
Updated guidance and temporarily removed JM Specialities Medium Phantom order form.