See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/312050027

The effect of quality control on the function of magnetic resonance imaging

(MRI), using American College of Radiology (ACR) phantom

Article  in  Egyptian Journal of Radiology and Nuclear Medicine · December 2016

DOI: 10.1016/j.ejrnm.2016.12.003

CITATIONS READS

3 461

5 authors, including:

H.M. Etman Ayman Mokhtar Said

1 PUBLICATION   3 CITATIONS   
Urology and Nephrology Center
10 PUBLICATIONS   11 CITATIONS   
SEE PROFILE
SEE PROFILE

Moustafa Tawfik Ahmed Tarek A El-Diasty

Mansoura University Urology and Nephrology Center
37 PUBLICATIONS   238 CITATIONS    182 PUBLICATIONS   2,729 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Big Data in Acute Renal Rejection View project

Modeling of the TSDC technique View project

All content following this page was uploaded by Tarek A El-Diasty on 23 January 2017.

The user has requested enhancement of the downloaded file.

 The Egyptian Journal of Radiology and Nuclear Medicine xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

The Egyptian Journal of Radiology and Nuclear Medicine

journal homepage: www.sciencedirect.com/locate/ejrnm

Original Article

The effect of quality control on the function of magnetic resonance

imaging (MRI), using American College of Radiology (ACR) phantom
H.M. Etman a, A. Mokhtar b, M.I. Abd-Elhamid a, M.T. Ahmed a,⇑, T. El-Diasty b
a
Physica Department, Faculty of Science, Mansoura University, Mansoura, Egypt
b
Radiology Department, Urology and Nephrology Center, Mansoura University, Mansoura, Egypt

a r t i c l e i n f o a b s t r a c t

Article history: Objective: To study image quality of MRI scanner using the American College of Radiology (ACR) phan-
Received 8 October 2016 tom.
Accepted 10 December 2016 Material and methods: Image quality of 1.5 T MRI scanner was tested using ACR phantom. A standard head
Available online xxxx
coil with standard restraints was used to fix the phantom head position. The phantom included seven
modules for measuring MRI scanner image quality. MRI images for each module were analyzed.
Keywords: Results: For the Geometric accuracy test the inside length of the phantom was 146 mm and the inside
Magnetic resonance imaging
diameter was 189.4 mm. For the High-contrast spatial resolution, individual small bright spots on the
ACR MRI phantom
MRI image quality
image were three pairs of hole arrays and were distinguishable. For the slice thickness test, the top signal
Quality assurance ramp length was 54.4 mm and bottom signal ramp length was 54.8 mm so the slice thickness was
5.46 mm. For the slice position accuracy, the bar length differences of intensity uniformity a value, called
percent integral uniformity (PIU), was 96.15%. For the percent-signal ghosting, ghosting ratio was 0.0002.
For the low contrast object detectability, the sum of numbers of complete spokes scored was 30 spokes.
Conclusions: Image quality tests were very important in acceptance of any MRI scanner after installation
and during maintenance. Using ACR phantom, these tests approve that the image parameters are
acceptable.
Ó 2016 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by Elsevier This
is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/
4.0/).

1. Introduction So, there are no enough national or international guidelines for

MRI quality assurance in comparison to imaging methods which
Magnetic Resonance Imaging (MRI) is an imaging modality that using ionizing radiation. American College of Radiology (ACR) has
does not use ionizing radiation to create diagnostic useful images. an accreditation program which includes a standardized image
It produces high quality images of the inside of the human body quality measurement protocol and a phantom [1].
[1]. Seven necessary estimation of MR image quality are usually
NMR was found at the same time by two physicists, Felix Bloch included as follows: geometric accuracy, high contrast spatial res-
in addition to Edward Mills Purcell, soon after the end of World olution, slices thickness accuracy, slice position accuracy, image
War II. They got the Nobel Prize in Physics in 1952 for this disclo- intensity uniformity, percent signal ghosting, and low contrast
sure. The use of NMR to produce 2D images was accomplished by object detectability [2].
Paul Lauterbur, imaging water and Sir Peter Mansfield who imaged Different scanners images should fulfill the same technical and
fingers of a research student, Dr. Andrew Maudsley in 1976. Maud- clinical standards that make regular quality assurance (QA) essen-
sley continues to make a significant contribution to MRI R&D. Ray- tial to MRI scanners. That is because there are limited national or
mond Damadian obtained human images a year later in 1977. international guidelines for MRI quality assurance, but substantial
Lauterbur and Mansfield received the Nobel Prize in Physiology work has been done.
or Medicine in 2003 for their development of MRI [1]. Our objective was to assess image quality using the ACR accred-
itation phantom.

Peer review under responsibility of The Egyptian Society of Radiology and Nuclear
Medicine.
⇑ Corresponding author.
E-mail address: [email protected] (M.T. Ahmed).

http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
0378-603X/Ó 2016 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by Elsevier
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article in press as: Etman HM et al. . Egypt J Radiol Nucl Med (2016), http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
2 H.M. Etman et al. / The Egyptian Journal of Radiology and Nuclear Medicine xxx (2016) xxx–xxx

2. Materials and methods Table 1

Parameters of the Scan Protocole.

2.1. Study design Study ACR sagittal ACR axial ACR axial T2 double-
locator T1 echo
This prospective study was approved by Institutional Research Pulse sequence Spin echo Spin echo Spin echo
Board (IRB). TR, ms 200 500 2000
TE, ms 20 20 50
The study was done on 18/7/2015 and repeated on 10/11/2015
FOV, cm 25 25 25
in the Urology and Nerphology Center in Mansoura University Number of slices 1 11 11
using Large ACR Phantom that fit into the head coil using a1.5 T Slice thickness, 20 5 5
Magnetic Resonance scanner. mm
Slice gap, mm – 5 5
NEX 1 1 1
Matrix 256
256 256
256 256
256
Scan time (min: 0:56 2:16 8:56
sec)
2.2. MRI scanner

The MRI scanner was 1.5 T (Signa Horizon LX Echo Echospeed,

General Electric Healthcare, Milwaukee, WI, USA). 2.4. Scanning protocol

The phantom was setting that the words ‘‘Nose” and ‘‘Chin”
means the nose and chin were located in a standard head study.
The middle of the phantom (the dark notch on the phantom) was
2.3. Large American College of Radiology (ACR) phantom
positioned in the middle of the head coil and the isocenter of the
scanner was aligned by the indicator light [4].
The ACR MRI phantom is constructed of acrylic plastic hollow
cylinder closed at both ends. The dimension of the phantom is
148 mm in length and 190 mm in diameter. It is filled with a nickel
chloride solution of about 10 mM NiCl2 and sodium chloride solu- 2.5. Image analysis
tion of about 75 mm NaCl. The words ‘‘CHIN” and ‘‘NOSE” is etched
outside the phantom to orienting phantom for scanning, as if it was The ACR accreditation program of the phantom portion was
a head [2]. As illustrated in Fig. 1. consists of seven tests [4].
The MRI phantom portion accreditation program required the
conquest of a sagittal localizer and images of four axial series.
Within the phantom there are eleven slice locations was conquest 2.5.1. Geometric accuracy
in each of the four axial series. The scanner’s head coil was used for This test assessed the accuracy of length image which the image
acquired these images. represent [5].
The sagittal laser alignment was checked to confirm that it was This test were measure as follow [2]:
match the line running along the phantom surface ‘‘nose” [2].
For the quality assurance, three different scans were made with 1. The localizer was displayed and the end-to-end length of the
parameters, specified by the ACR accreditation program. These phantom was measured [6]. Fig. 2a.
scan protocols and parameters are shown in Table 1 [3]. 2. Slice one of the ACR T1 series was displayed and the phantom
After the locator eleven slices follow, starting with slice 1 and diameter in two directions was measured: left-to-right and
ending with slice 11.The evaluations of images were performed top-to-bottom Fig. 2b.
as instructed by the ACR. The evaluation was done with Digital 3. Slice five of the ACR T1 series was displayed and the diameter in
Imaging and Communications in Medicine (DICOM) viewer, four directions of the phantom was measured: left-to-right,
because it is one of the non-commercial suitable DICOM image top-to-bottom, and both diagonals Fig. 2c [7].
viewer software, recommended by the ACR [1].

Fig. 2a. The sagittal localizer of the end-to-end length measurement illustrated
Fig. 1. Photograph of ACR MRI phantom. (arrow).

Please cite this article in press as: Etman HM et al. . Egypt J Radiol Nucl Med (2016), http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
 H.M. Etman et al. / The Egyptian Journal of Radiology and Nuclear Medicine xxx (2016) xxx–xxx 3

2. We started with the most hole arrays in the left most pair, the
pair which have the largest hole size contrast, 1.1 mm and the
rows of holes in the UL array was looked (If all 4 holes in any
single row are distinguishable from one another, the image
was scored as resolved) [7].

2.5.3. Slice thickness accuracy

In this test we measured the two signal ramps lengths in slice
one. The structure of ramps called the slice thickness insert [5].
The measurement of the signal ramps length in slice 1 in each
ACR series is determined by the following method:

1. Slice one was displayed, and the image was magnified until the
slice thickness insert was fully visible on the screen and the dis-
play level was adjusted so that the signal ramps were
visualized.
2. A rectangular Region of Interest (ROI) was placed at the middle
Fig. 2b. The diameter of slice 1 measurements illustrated (arrows). of each signal ramps, Fig. 4. The mean signal values of the two
ROIs were noted, and the average of 2 values was calculated.
The result was a number approximating the mean signal in
the middle of the ramps.

The following formula was used to calculate the slice thickness

[7]:

Slice thickness ¼ 0:2
ðtop
bottomÞ=ðtop þ bottomÞ:

2.5.4. Slice position accuracy

This test determine the specific locations of slices to be accurate
using the localizer image as positional Ref. [5]
Measurements were taken for slices one and eleven of each two
ACR series. The following steps were done for each image slice [2]:

1. The image slice was displayed and the vertical bars of the
crossed wedges were kept the magnified image within the dis-
played portion.
Fig. 2c. Slice 5 with diameter measurements illustrated (arrows). 2. Adjustment of display window was done so the vertical bars
ends were well defined.
3. The bars length difference between the right and left is mea-
2.5.2. High-contrast spatial resolution sured using the screen length measurement tool. The arrows
This test determine the ability of scanner to resolve small illustrated the length in Figs. 5a and 5b.
objects [5].
In this test, in slice one we evaluate the resolution of each of the
2 ACR axial series. The following method was repeated for each of
two series: 2.5.5. Image intensity uniformity
This test assessed a solution only region of the phantom. This
1. The image of slice one was displayed and the image was mag- was assessed by determining and comparing the high and low
nified until the resolution insert of the displayed image was vis- intensity levels within the phantom region of slice 7 [5].
ible. This is illustrated in Fig. 3. The following procedure were made for the 2ACR series accord-
ing to as follow [2]:

Fig. 3. Magnified portion of slice 1 displayed appropriately for visually assessing high contrast resolution (appearance of well-resolved holes).

Please cite this article in press as: Etman HM et al. . Egypt J Radiol Nucl Med (2016), http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
4 H.M. Etman et al. / The Egyptian Journal of Radiology and Nuclear Medicine xxx (2016) xxx–xxx

Fig. 4. The slice thickness insert of slice 1 and ROIs of signal ramps placed for measuring average signal in the ramps.

Fig. 5a. Images of slice 1 with the pairs of vertical bars from the 45° crossed wedges indicated.

Fig. 5b. The pairs of vertical bars from the 45° crossed wedges of slice 11 indicated.

Please cite this article in press as: Etman HM et al. . Egypt J Radiol Nucl Med (2016), http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
 H.M. Etman et al. / The Egyptian Journal of Radiology and Nuclear Medicine xxx (2016) xxx–xxx 5

Fig. 6. The image illustrates the size and the large of slice 7.

1. Slice location seven was displayed. A big, circular region-of- 1. Slice seven was displayed. A big, circular ROI on the image was
interest (ROI) on the image was placed, Fig. 6. This ROI have placed as shown in Fig. 7. This ROI had an area of between
an area of between 19,500 mm2 and 20,500 mm2. The ROI 195 cm2 and 205 cm2 (19,500–20,500 mm2). The mean pixel
defined the boundary of the measured region of the image value for this ROI was recorded.
uniformity. 2. Elliptical ROIs was placed along the four edges of the field of
2. The level slowly was raised until about 1 cm2 region of dark pix- view, as shown in Fig. 7. The mean pixel value for each ROI
els developed inside the ROI. The large ROI was the region of was recorded.
lowest signal in. A 1 cm2 circular ROI was placed on the low-
signal region and then the mean pixel value was recorded and The ghosting ratio was calculated by using this following:
the low signal value was measured.
Ghosting ratio ¼ jððtop þ btmÞ  ðleft þ rightÞÞ=ð2 xðlarge ROIÞÞj
The Percent Integeral Uniformity (PIU) was produced by com- where top, bottom, right, left, and big ROI are the mean pixel values
bined the measured high – low signal value of the ACR series. for the ROIs of the same names.
The PIU was calculated by this formula:
2.5.7. Low-contrast object detectability
In this test the extent the discernible low contrast object in the
PIU ¼ 100
ð1  ðhigh  lowÞ=ðhigh þ lowÞÞ ½8:
images was assessed [5].
The low-contrast disks is a holes drilled in thin sheets of plastic
where high is the measured high-signal value and low is the mea-
mounted in the phantom at four slices locations.
sured low-signal value.
The number of complete spokes was counted in this test in each
of the four slices [2].
2.5.6. Percent-signal ghosting The number of complete spokes was scored according to the fol-
This test assessed image ghosting level. Ghosting was an artifact lowing procedure:
in which a faint copy of the imaged object appears on the image
and is displaced from its true location [5]. 1. The slice to be scored was displayed. We started with eleventh
In this test, the following procedures for the ACR T1 series were slice, which had the highest contrast objects.
made on slice seven.

Fig. 7. The image illustrates the ROI placement for the percent signal ghosting of slice 7.

Please cite this article in press as: Etman HM et al. . Egypt J Radiol Nucl Med (2016), http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
6 H.M. Etman et al. / The Egyptian Journal of Radiology and Nuclear Medicine xxx (2016) xxx–xxx

2. The numbers of complete spokes was counted. We count the 3.5. Image intensity uniformity
largest diameter disks of complete spoke firstly and then the
score was recorded. As shown in Fig. 8a–8d) The percent integral uniformity (PIU), was calculated from the
slice number 7 of ACR T1 (Fig. 6) and ACR T2 sequences. The PIU
Note (A spoke was complete only if all three of its disks were was 96.2%.
discernible. Complete spokes was counted, not individual disks).
3.6. Percent-signal ghosting
2.6. Data analysis
Ghosting ratio is calculated from the seventh slice of ACR T1
The obtained image quality parameters were compared with scans using one big and four small regions of interest.
the tolerance values of ACR. The evaluation was done with DICOM The ghosting ratio was calculated 0.0002 which was less than or
viewer, because it was one of the non-commercial suitable DICOM equal to 0/025. This indicated that scanners passed the test of per-
image viewer software. cent signal ghosting.

3.7. Low-contrast object detectability

3. Results

Disks have been seen in the eighth slice to eleventh slice for T1
All ACR MRI quality assurance tests were performed success-
sequence. The sum of spokes was 30.
fully. Analysis of MRI images showed that the parameters of image
quality were within ACR guidelines for the scanned protocol.
4. Discussion
3.1. Geometric accuracy
The essential results in our study were the Geometric accuracy
test seven measurements were made which was accepted with the
The inside length of the phantom was 146 mm and the inside
tolerance value. For the High-contrast spatial resolution test we
diameter was 189.4 mm which was accepted with the tolerance
can in visual evaluation differentiate holes with the diameter of
value.
1.0 mm at least. For the Slice thickness accuracy test the two spe-
cial signal ramps lengths were measured. For the Slice position
3.2. High-contrast spatial resolution accuracy test the length difference of special bars in two slices
was measured which was considered within tolerance value. For
The three whole array pairs were seen on the first ACR T1 slice the Image intensity uniformity the PIU was 96.2% which is more
in Fig. 3, where they were located at the bottom side of the slice. In than or equal to 87.5% with field strengths less than 3 T for the
the images the hole array pair appeared like two squares linked MRI systems. For the Percent-signal ghosting the ghosting ratio
together through one corner and filled with small bright dots. was calculated 0.0002 which was less than or equal to 0/025. For
the Low-contrast object detectability test the sum of spoke was
30, this means the low-contrast objects images were discernible.
3.3. Slice thickness accuracy
Quality Assurance (QA) of medical imaging systems is impor-
tant and necessary in a modern hospital. Using ACR MRI phantom
The prescribed slice thickness, 5.0 mm, was compared to the
for quality assurance purpose has many advantages. With ACR MRI
measured slice thickness which was 5.46 mm and the measured
phantom, it is possible to test many image quality parameters that
slice thickness was 5.0 ± 0.7 mm.
manufacturers do not test on their regular check-ups.
The ACR MRI phantom is independent of manufacturers, provid-
3.4. Slice position accuracy ing the possibility to be used in different MRI scanners and to com-
pare these test results to one another. In addition, the ACR has an
The bar length difference was 4.7 mm which was considered internationally recognized status as a developer of guidelines and
within the tolerance value. standards for radiology devices and QA procedures. The image

Fig. 8a. The slice 8 image has the circle of low contrast objects for the low-contrast object detectability test. (This image has 3 spokes.)

Please cite this article in press as: Etman HM et al. . Egypt J Radiol Nucl Med (2016), http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
 H.M. Etman et al. / The Egyptian Journal of Radiology and Nuclear Medicine xxx (2016) xxx–xxx 7

Fig. 8b. The slice 9 image has the circle of low contrast objects for the low-contrast object detectability test. (The image has 8 spokes.)

Fig. 8c. The slice 10 image has the circle of low contrast objects for the low-contrast object detectability test. (This image has 8 spokes.)

Fig. 8d. The slice 11 image has the circle of low contrast objects for the low-contrast object detectability test. (This image has 10 spokes.)

acquisition procedure is fast and with one phantom it is possible to the adjustments had been made by service engineers, all the scan-
assess seven different important quality parameters [3]. ners passed the QA tests. It was stated that the proposed ACR MRI
Similar studies [3,6,9] with the ACR MRI phantom have been QA protocol provides a simple and comprehensive assessment of
made earlier. In Kaljuste and Nigul study [3] using the ACR MRI the performance of an MRI scanner. In Ihalainen et al. study [9]
phantom for quality assurance purpose is a feasible way to detect two tests were made and eleven MRI systems were tested. On
if the tested MRI scanners fulfill the standards of a well-functioning the first measurements, 91% and on the second measurements
MRI scanner. Even though manufacture specific service programs 73% of systems passed the tests. It was stated that the image acqui-
are used, it is important to use standard phantoms in MRI quality sition procedure of the ACR tests was fast and practical and the
assurance to ensure that minimum image quality level is equal method proved feasible for quality assurance in the multi-unit
to all MRI scanners in practice. In Chen et al. study [6] half of the imaging center, where the tests were made.
four scanners tested passed the QA tests at the first time and after

Please cite this article in press as: Etman HM et al. . Egypt J Radiol Nucl Med (2016), http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
8 H.M. Etman et al. / The Egyptian Journal of Radiology and Nuclear Medicine xxx (2016) xxx–xxx

For future tests, special structures should be designed for every Conflict of interest
used head coil to position the phantom accurately to the isocenter
of the coil because the ACR MRI phantom did not fit inside some The authors declare that there is no conflict of interest.
head coils so using different coils can influence some quality assur-
ance tests of the MRI scanner. References
For better quality assurance, signal-to-noise ratio assessment
should be added to the seven assessments because of its funda- [1] Haacke EM, Brown RW, Thompson MR, Venkatesan R. Magnetic resonance
imaging: physical principles and sequence design, vol. 82. New York: Wiley-
mental position in the MRI quality assurance. Liss; 1999.
[2] American College of Radiology. Phantom test guidance for the ACR MRI
accreditation program. Reston, Va: ACR; 1998.
[3] Kaljuste D, Nigul M. Evaluation of the ACR MRI phantom for quality assurance
tests of 1.5 T MRI scanners in Estonian hospitals. Proc Estonian Acad Sci 2014;63
5. Conclusion (3):240.
[4] American College of Radiology. Site scanning instructions for use of the MR
phantom for the ACR MRI accreditation program. Reston, VA: American College
To the best of our knowledge, this is the first study that was car- of Radiology; 2000.
ried out in Egypt. [5] Clarke GD. Overview of the ACR MRI accreditation phantom. Dallas: University
of Texas Southwestern Medical Center; 2004.
The MRI scanner was tested with the parameters stated by the [6] Chen CC, Wan YL, Wai YY, Liu HL. Quality assurance of clinical MRI scanners
ACR MRI accreditation program. Of the tested scanner, all the ACR using ACR MRI phantom: preliminary results. J. Digit. Imaging 2004;17
MRI quality assurance tests were performed successfully. (4):279–84.
[7] Shu Y, Gorny KR, Felmlee JP, Pooley RA, Edmonson HA. Practical considerations
All the previous studies were in Europe only, so we recommend
for ACR MRI accreditation. J Am College Radiol 2014;11(1):94–6.
that image quality test for MRI must be performed in other Egyp- [8] Al Mohammad B. Quantitative MRI quality control using the ACR phantom;
tian and African centers, because image quality tests are very 2012.
important tests in acceptance of any MRI scanner after installation [9] Ihalainen TM, Lonnroth NT, Peltonen JI, Uusi-Simola JK, Timonen MH, Kuusela L.
J, et al. MRI quality assurance using the ACR phantom in a multi-unit imaging
and during maintenance to approve that the image parameters are center. 50; 2011. p. 966–972.
accepted by using ACR phantom.

Please cite this article in press as: Etman HM et al. . Egypt J Radiol Nucl Med (2016), http://dx.doi.org/10.1016/j.ejrnm.2016.12.003
View publication stats