King’s Research Portal

DOI:10.1007/s12350-015-0323-0

Document VersionPublisher's PDF, also known as Version of record

Link to publication record in King's Research Portal

Citation for published version (APA):Polycarpou, I., Chrysanthou-Baustert, I., Demetriadou, O., Parpottas, Y., Panagidis, C., Marsden, P. K., &Livieratos, L. (2015). Impact of respiratory motion correction on SPECT myocardial perfusion imaging using amechanically moving phantom assembly with variable cardiac defects. Journal of Nuclear Cardiology.10.1007/s12350-015-0323-0

Citing this paperPlease note that where the full-text provided on King's Research Portal is the Author Accepted Manuscript or Post-Print version this maydiffer from the final Published version. If citing, it is advised that you check and use the publisher's definitive version for pagination,volume/issue, and date of publication details. And where the final published version is provided on the Research Portal, if citing you areagain advised to check the publisher's website for any subsequent corrections.

General rightsCopyright and moral rights for the publications made accessible in the Research Portal are retained by the authors and/or other copyrightowners and it is a condition of accessing publications that users recognize and abide by the legal requirements associated with these rights.

•Users may download and print one copy of any publication from the Research Portal for the purpose of private study or research.•You may not further distribute the material or use it for any profit-making activity or commercial gain•You may freely distribute the URL identifying the publication in the Research Portal

Take down policyIf you believe that this document breaches copyright please contact librarypure@kcl.ac.uk providing details, and we will remove access tothe work immediately and investigate your claim.

Download date: 18. Feb. 2017

ORIGINAL ARTICLE

Impact of respiratory motion correction onSPECT myocardial perfusion imaging using amechanically moving phantom assembly withvariable cardiac defects

Irene Polycarpou, PhD,a,b Isabelle Chrysanthou-Baustert, PhD,c

Ourania Demetriadou, MD,d Yiannis Parpottas, PhD,c,e Christoforos Panagidis,

MD,f Paul K. Marsden, PhD,a and Lefteris Livieratos, PhDa,g

a Division of Imaging Sciences and Biomedical Engineering, King’s College London, London,

United Kingdomb Department of Health Sciences, European University Cyprus, Nicosia, Cyrpusc Frederick Research Center, Nicosia, Cyprusd Department of Nuclear Medicine, Limassol General Hospital, Limassol, Cypruse General Department (Physics-Mathematics), Frederick University, Nicosia, Cyprusf Department of Nuclear Medicine, Nicosia General Hospital, Nicosia, Cyprusg Guy’s and St Thomas’ Hospitals NHS Foundation Trust, London, United Kingdom

Received Jul 27, 2015; accepted Oct 26, 2015

doi:10.1007/s12350-015-0323-0

Background. The aim of this study was to determine the impact of respiratory motioncorrection on SPECT MPI and on defect detection using a phantom assembly.

Methods. SPECT/CT data were acquired using an anthropomorphic phantom withinflatable lungs and with an ECG beating and moving cardiac compartment. The heart motionfollowed the respiratory pattern in the cranio-caudal direction to simulate normal or deepbreathing. Small or large transmural defects were inserted into the myocardial wall of the leftventricle. SPECT/CT images were acquired for each of the four respiratory phases, from exhaleto inhale. A respiratory motion correction was applied using an image-based method withtransformation parameters derived from the SPECT data by a non-rigid registration algo-rithm. A report on defect detection from two physicians and a quantitative analysis on MPIdata were performed before and after applying motion correction.

Results. Respiratory motion correction eliminated artifacts present in the images, resultingin a uniform uptake and reduction of motion blurring, especially in the inferior and anteriorregions of the LV myocardial walls. The physicians’ report after motion correction showed thatimages were corrected for motion.

Conclusions. A combination of motion correction with attenuation correction reducesartifacts in SPECT MPI. AC-SPECT images with and without motion correction should besimultaneously inspected to report on small defects. (J Nucl Cardiol 2015)

Key Words: SPECT/CT Æ myocardial perfusion defects Æ motion correction Æ cranio-caudalcardiac motion

Reprint requests: Irene Polycarpou, PhD, Division of Imaging Sciences

and Biomedical Engineering, King’s College London, London,

United Kingdom; polycarpou@kcl.ac.uk

1071-3581/$34.00

Copyright � 2015 American Society of Nuclear Cardiology.

AbbreviationsSPECT Single-photon emission computed

tomography

CT Computed tomography

AC Attenuation correction

MPI Myocardial perfusion imaging

PLC Programmable logic controller

MC Motion correction

LV Left ventricle

INTRODUCTION

In clinical SPECT studies, respiratory motion limits

the performance of MPI studies. Heart motion induced

by respiration affects SPECT acquisition in two ways.1

First, spatial misalignment between sequentially

acquired SPECT and CT may affect attenuation correc-

tion of the emission data since SPECT images are

acquired for substantially longer time than the CT

images. Various studies have shown that such a

misalignment may cause image artifacts and under- or

over-estimation of tracer activity.2 Furthermore, respi-

ration may lead to image blurring along the direction of

motion, primarily in the cranio-caudal direction, due to

the fact that SPECT data are averaged over many

respiratory cycles. This can result in misleading dis-

placement of uptake affecting the perceived regional

localization in the myocardial walls.3,4 Motion-induced

artifacts can be a source of false-positive findings and

may be interpreted as ischemia affecting the diagnosis of

coronary artery disease.5

A non-uniform blurring of MPI due to motion

induced by respiration4 may affect the ability to detect

defects.6 The respiratory motion of the heart has been

previously investigated using computational4 and phys-

ical phantoms.3,8 It was simulated on physical phantoms

with a static cardiac compartment8 being on an oscil-

lating plate or with static cardiac compartments3,9,10

inserted within torso phantoms and then being on

oscillating plates. The effects of respiratory motion on

MPI non-uniformity and on the resulting false-positive

rates have also been reported.3,4 A common method to

compensate for motion effects is data gating which

divides the motion cycle into a number of gates based on

either time or amplitude.7,11 Further, the respiratory

motion of the heart on physical phantoms8,10 was

corrected using motion correction algorithms. However,

limited studies, only with computational phantoms6

using the channelized Hotelling observer algorithm,

reported on the effects of respiratory motion on myocar-

dial defect detection.

In this study, the impact of respiratory motion

correction on MPI and on defect detection was inves-

tigated for heart motion amplitudes of normal and deep

breathing. An anthropomorphic phantom, with inflat-

able lungs, and an ECG beating and moving cardiac

compartment, was used to simulate the heart respiration

motion. Motion correction was applied using an image-

based method with transformation parameters derived

from the SPECT data by a non-rigid registration

algorithm. SPECT/CT images were acquired for each

of the four respiratory gates, from exhale to inhale, in

normal and deep breathing modes of the phantom

assembly. Defects could be inserted within the myocar-

dial wall of the LV. Two physicians reported on defect

present or absent before and after correcting for motion

of the AC-SPECT images. A comparison of the quan-

titative analysis and physicians’ reports before and after

motion correction is presented.

MATERIALS AND METHODS

Phantom Assembly

A cardiac phantom of an ECG beating and moving left

ventricle with variable defects, and a respiratory phantom of

inflatable lungs, constructed at Frederick Research Center,

were inserted into an anthropomorphic thorax phantom12 and

used to simulate the heart motion during respiration using

routine imaging conditions.

The cardiac phantom consisted of an inner–outer mem-

brane system where the inner membrane represented an LV and

the closed cavity within the inner and outer membranes

represented the myocardial wall of the LV. Each 2-mm thick

ellipsoidal membrane was made of transparent silicone with a

density of 1.06 g/cm2. The end systolic volume was 42 mL and

the LV myocardial wall volume was 90 mL. Diastole and

systole could be achieved by pumping water into and out of the

inner membrane using a stepper motor controlled by the PLC

and a piston. Since the myocardial wall cavity was filled with

water which is non-compressive, the pressure was transferred

from the inner to the outer membrane (Figure 1A). The beating

rate (60 beats/minute) and the ejection fraction (45%) followed

the Wigger diagram. A 5% deviation was found using the

Xeleris Myovation (GE Healthcare) compared to the set

ejection fraction. Radionuclide could be injected within the

myocardial wall cavity. Defects could be glued on the outer part

of the inner membrane (Figure 2A). The ECG signal from a

simulator was simultaneously propagated to the PLC to initiate

diastole–systole and to the gating signal circuit which fed into

the gamma-camera system to trigger gating of data acquisition.

The respiratory phantom consisted of human-sized sili-

cone lungs (Figure 2B). The left lung air volume was 510 mL

and the right lung air volume was 565 mL. Each lung was

covered with a 1-mm-thick thermoplastic structures forming

the lung shape at the inhalation phase allowing inflation in an

anatomically correct manner. The breathing cycle and tidal

Polycarpou et al Journal of Nuclear Cardiology�SPECT myocardial perfusion imaging

volume followed a sinusoidal exponential pattern.13 Two

electro-proportional valves, controlled by the respiratory pat-

tern equation in the PLC, regulated the air pressure with

respect to time, entering in the lungs for inhalation and

escaping from the lungs for exhalation, respectively (Fig-

ure 1B). This was achieved using two small pumps with

suction and blowing ports that could be used as vacuum pumps

and small compressors. The tidal volume was 450 and 550 mL

for the normal and deep breathing, respectively. In both cases,

the breathing cycle duration was 6 seconds. A 2% deviation

was found using OsiriX14 compared to the set lung volumes.

The cardiac compartment could follow the respiratory

pattern in the cranio-caudal direction at the level of the

diaphragm. The mechanical system for this motion, a

controlled motor by the PLC and a screwed rod, was attached

to the outer base of the anthropomorphic phantom (Figures 1C,

3). The beating cardiac compartment was set to perform a

maximum displacement of 1.5 and 2.7 cm for the normal and

deep breathing conditions, respectively.15 All three motions

could be independently or simultaneously controlled using a

PLC.

The respiratory pattern was divided into four isochronous

phases (from exhale to inhale). Thus, the breathing lungs and

the moving heart could be mechanically stopped at any

respiratory phase, as shown in Table 1, while the heart was

continuously beating.

The Alderson tissue-equivalent anthropomorphic phan-

tom for nuclear medicine (Radiology Support Devices, Inc.)

Figure 1. Phantom assembly motions: (A) systole and diastole of the cardiac phantom,(B) breathing of the respiratory phantom, and (C) motion of the cardiac phantom during respiration.

Figure 2. Photos of (A) inner and outer membranes of the cardiac compartment with a defect,(B) cardiac compartment in systole and lung compartment at the exhale phase, and (C) cardiaccompartment with defect in diastole and lung compartment at the inhale phase.

Journal of Nuclear Cardiology� Polycarpou et al

SPECT myocardial perfusion imaging

was used to simulate an average male and a large male by

adding a thorax overlay, and a large female by adding breasts

on the thorax overlay (Figure 3). The cardiac and lung

compartments were anatomically hold within the thorax using

a shaped thin solid-water slab. The thorax was hermetically

closed and the rest of the cavity was filled with water to

simulate the soft tissue.

Data Acquisition

Acquisitions were performed for three body types with no

cardiac defects, with two small (0.5 cm thickness, 1 9 1 cm in

extent) and with two large (1 cm thickness, 1 9 1 cm in

extent) transmural defects. The defects were positioned at the

anterior and inferior regions of the LV myocardial wall

(Table 2). An activity of 15.5 MBq of Tc99m was injected into

the myocardium wall of the LV based on an estimate of 1.2%

uptake of a clinically relevant administered activity.16

SPECT data were acquired in a GE Millennium VG

gamma-camera with a single-slice Hawkeye CT at Nicosia

General Hospital. Acquisitions were performed using a routine

clinical protocol with Low-Energy High-Resolution collima-

tors in 90� (L-mode) orientation. SPECT data were acquired in

60 projections over 180� of rotation on 64 9 64 matrix

following the default clinical protocol at 20 seconds per

projection. Data were acquired at the required respiratory

phase while the heart was continuously beating. A 20% energy

window was centered over the 140 keV photopeak of Tc99m.

CT images were acquired in the heart region with 10 mm slice

thickness and a matrix size of 128 9 128. SPECT and CT

images were acquired in each of the four respiratory phases for

the normal and deep breathing, respectively (Table 2).

SPECT data were reconstructed using the ordered-subset

expectation-maximization algorithm with three iterations and

16 subsets, as commonly used in clinical practice. The image

matrix size was 64 9 64 9 19 with a voxel size of

6.9 9 6.9 9 6.9 mm. The Butterworth filter (cutoff: 0.52

cycles/cm, power: 5) was applied to the reconstructed images.

Attenuation correction was applied based on CT-derived

attenuation maps. SPECT data were generated with and

without motion correction. All data were re-oriented to form

short-axis images along the long axis of the LV processed by

the same operator in the same way for all reconstructions with

and without motion correction.

Figure 3. Photos of (A) average male, (B) large male, and (C) large female body types of theanthropomorphic phantom.

Table 1. Lung volumes and cardiac displacement at each of the four respiratory phases, from exhale(phase-1) to inhale (phase-4), for the normal and deep breathing modes

Phase

Normal breath Deep breath

Volume (mL) Displacement (mm) Volume (mL) Displacement (mm)

1 1075 0 1075 0

2 1201 5 1201 5

3 1373 10 1433 20

4 1538 15 1635 27

Table 2. SPECT/CT acquisitions for the threebody types, defect types, and breathing modes

Body type Defects Breathing mode

Average male None Static, normal, deep

Two small Static, normal, deep

Two large Static, normal, deep

Large male None Static, normal

Two small Static, normal, deep

Two large Static, normal, deep

Large female None Static, normal, deep

Two large Static, normal, deep

Polycarpou et al Journal of Nuclear Cardiology�SPECT myocardial perfusion imaging

Motion Correction

Motion correction was applied using an image-based

procedure. A non-linear registration algorithm17 was applied

to register each of the reconstructed SPECT respiratory

phases (gates) to the reference gate (lungs at the exhale phase

and no heart motion) allowing for possible non-rigid motion

patterns in the clinical setting in line with previous studies.18

In this way, transformation fields (i.e., displacement vectors

between the reference gate and each of the other gates)

describing the motion were obtained. The registration algo-

rithm is fully automated voxelwise which estimates non-rigid

motion by a combination of multiple local affine components

based on an adaptive hierarchical structure.17 After deriving

the motion fields, each gate was transformed to the reference

gate and these were averaged to create the motion-corrected

image.

A motion-corrected image was generated from the four

SPECT/CT acquisitions obtained when the phantom assembly

was mechanically stopped in each of the four respiratory

phases, for each breathing mode, defect type, and body type

(Table 2). In total, fifteen motion-corrected AC-SPECT

images (8 for the normal and 7 for the deep breathing modes)

were generated. These images were compared with the

corresponding AC-SPECT images (SPECT with breathing

lungs and moving heart, and AC with CT at the reference

gate). Figure 4 shows an example of co-registration of SPECT

with CT.

Image Assessment

The motion-corrected AC-SPECT images were qualita-

tively and quantitatively compared to the AC-SPECT images

to assess the effectiveness of motion correction in reducing

motion artifacts. For the quantitative comparison, the percent-

age increase of the myocardial perfusion uptake values after

applying motion correction was calculated for each of the 17

segments of the polar maps.

Two experienced nuclear medicine physicians reported on

images without defects and with one or two defects. The

physicians did not know about the phantom parameters or

image correction (AC, MC). The physicians reported on the

defect absence or presence. The physicians’ report was correct

if the defect position was correctly identified. The results from

the report on each defect were evaluated as True Positive (TP),

True Negative (TN), False Positive (FP), and False Negative

(FN).

RESULTS AND DISCUSSION

Figure 5 shows SPECT images of the four respira-

tory phases obtained for the large female with no defects

performing normal breathing. Figures 6 and 7 show AC-

SPECT images without motion and performing normal

and deep breathing before and after MC, obtained for the

average male with no defects and the large female with

large defects, respectively. Motion-induced artifacts

cause seemingly reduced anterior and inferior wall

uptake introducing false-positive findings. The corre-

sponding polar maps of Figure 7 are shown in Figure 8

together with the percentage increase of the myocardial

perfusion uptake for each LV segment after applying

motion correction. A visual inspection of the AC-

SPECT images shows that motion due to respiration

causes blurring and defect smearing implying a much

extensive disease compared to the motion-corrected

images. In particular, heart motion due to respiration

causes under-estimation of the tracer uptake mainly in

the inferior and anterior regions of the LV myocardial

wall. This was also reported in previous studies.3,4,7 The

apical artifact observed in some figures is due to a solid-

water square cross (0.5 cm side, 1 mm thickness)

attached at the apex of the cardiac compartment to

prevent the inner membrane touching the outer mem-

brane during diastole.

Motion correction substantially reduces the motion-

induced artifacts due to respiration observed in both

SPECT slices and polar maps. In particular, images after

motion correction result in more uniform uptake and

reduction of motion blurring, a thinner appearance of the

heart wall, especially in the inferior and anterior regions

of the LV myocardial wall. An increase of the uptake

values was also observed in the septal and lateral regions

after motion correction but not as severe as that

observed in the inferior and anterior regions. After

motion correction, the percentage increase of the MPI

uptake values in the anterior and inferior regions was

18% for normal and 35% for deep breathing, while the

corresponding increase in the apical, septal, and lateral

regions was 7% for normal and 9% for deep breathing.

The quantitative analysis shows that after motion cor-

rection, there is major improvement in the inferior and

anterior regions which are expected to be more affected

by motion.

Figures 9 and 10 show SPECT images with (A)

non-AC and non-MC, (B) AC and non-MC, (C) non-AC

and MC, and (D) AC and MC, for the average male with

small defects performing normal and deep breathing,

respectively. Motion correction does not recover the

uptake without applying attenuation correction. Apply-

ing only attenuation correction does not eliminate

blurring in the anterior and inferior regions due to

Figure 4. Example of transverse, coronal, and sagittal planesof the registered SPECT and CT for the average male with nodefects.

Journal of Nuclear Cardiology� Polycarpou et al

SPECT myocardial perfusion imaging

motion. The extent of improvement depends on the

amplitude of the heart motion.

Two physicians reported on the presence or absence

of cardiac defects in AC-SPECT images, with non-MC

and MC, for the normal and deep breathing modes. The

physicians’ results were considered to be in very good

agreement (Cohen’s Kappa coefficient: 0.86). Motion

led to FP due to artifacts. For example, when no motion

was present in phantoms with no defects, the physicians

reported no defects, which led to a TP evaluation.

However, when the phantoms with no defects performed

normal and deep breathing, the physicians reported false

Figure 5. SPECT slices of short, horizontal long, and vertical long axis, obtained for the largefemale with no defects, performing normal breathing: (A) gate-1; (B) gate-2; (C) gate-3; (D) gate-4.

Figure 6. AC-SPECT slices of short, horizontal long, and vertical long axis, obtained for theaverage male with no defects, without motion (A) and performing normal (B, C) and deep (D, E)breathing before (B, D) and after MC (C, E).

Polycarpou et al Journal of Nuclear Cardiology�SPECT myocardial perfusion imaging

defects in 4 out of 6 images and in 3 out of 4 images,

respectively, which led to an FP evaluation.

The 12 large defects were correctly reported in all

AC-SPECT images, that is, when the phantoms were in

static, normal, and deep breathing modes. Only 4 out of

the 8 small defects were reported in the AC-SPECT

images, when motion was not present (Table 3). How-

ever, when the phantoms performed normal and deep

breathing, all 8 small defects were reported in the AC-

SPECT images. This is because the defects were located

15%

24%

11%

4%

10%

9%

8%

3%

-5%

10%

1%

2%

5% 3%

7%

7% 9%

35%

32%

34%

13%

40%

45%

32%

5%

-4%

10%

5%

8%

3% 5%

12%

13% 6%

A B C

FD E

Figure 8. Quantitative analysis of the AC-SPECT polar maps obtained before (A, D) and after (B,E) applying MC and the percentage increase of the myocardial perfusion uptake (C, F) in each LVsegment after applying MC, for the large female with large defects performing normal (A, B, C)and deep breathing (D, E, F).

Figure 7. AC-SPECT slices of short, horizontal long, and vertical long axis, obtained for the largefemale with large defects, without motion (A) and performing normal (B, C) and deep (D, E)breathing before (B, D) and after MC (C, E).

Journal of Nuclear Cardiology� Polycarpou et al

SPECT myocardial perfusion imaging

Figure 9. Slices of short, horizontal long, and vertical long axis obtained for the average male withsmall defects performing a normal breathing: (A) non-AC and non-MC; (B) AC and non-MC; (C)Non-AC and MC; and (D) AC and MC.

Figure 10. Slices of short, horizontal long, and vertical long axis obtained for the average malewith small defects performing a deep breathing: (A) non-AC and non-MC; (B) AC and non-MC;(C) Non-AC and MC; and (D) AC and MC.

Polycarpou et al Journal of Nuclear Cardiology�SPECT myocardial perfusion imaging

in the anterior and inferior regions in which the uptake

values were mainly reduced due to motion-induced

artifacts. This was also reported in a similar study with

the NCAT phantoms.6

A combination of motion correction with attenua-

tion correction enhanced the reduction of artifacts.

Attenuation correction improved the uptake values in

the basal segments19, while motion correction improved

the uptake values in the anterior and inferior regions and

to a lesser extent in the lateral and septal regions.

Applying motion correction on AC-SPECT images, the

uptake values were increased in the anterior and inferior

regions, in which the small defects were located, with

the result 50% of the defects to be reported and

evaluated as FN, as in the case of static phantoms

(Table 3). Even though the images were corrected for

motion, to report on the presence or absence of small

defects, AC-SPECT images with and without motion

correction should be simultaneously inspected. The

study assumed the absence of misalignment of emission

and transmission (CT) data due to patient motion, and

therefore it represents an ideal case scenario with respect

to attenuation correction. The impact of attenuation

correction in the presence of motion-related misalign-

ment between the SPECT and CT has been highlighted

previously1, and the phantom of this study may be used

in future to investigate such effects.

NEW KNOWLEDGE GAINED

A combination of motion correction with attenua-

tion correction reduces artifacts in SPECT MPI. In AC-

SPECT images, small defects located on the anterior and

inferior regions of the myocardial wall of the left

ventricle become more visible due to motion-induced

artifacts. AC-SPECT images with and without motion

correction should be simultaneously inspected to be able

to report on small defects. The defects smeared by the

respiratory motion implying a much extensive disease

compared to the motion-corrected images leading to

misleading interpretation. Therefore, motion correction

is important for better diagnosis as it decreases the

smearing effect. Motion correction does not fully

recover the uptake without applying attenuation correc-

tion, provided that no misalignment between emission

and transmission data is present or has been corrected

for.

CONCLUSIONS

This study investigated the impact of respiratory

motion correction on AC-SPECT MPI and on defect

detection using a moving phantom assembly with

cardiac defects.

The inflatable lungs, and the ECG beating and

moving LV during respiration, simulated normal and

deep breathing. SPECT/CT data were acquired for each

of the four respiratory phases, from exhale to inhale, in

normal and deep breathing modes of the phantom

assembly. These SPECT/CT images were used to

generate the motion-corrected images. A non-linear

registration algorithm was applied to register each of the

four reconstructed respiratory phases to a reference gate

to obtain the transformation fields and generate motion-

corrected images. The quantitative analysis on AC-

SPECT images and the physicians’ report on defect

presence or absence, before and after applying motion

correction, were evaluated.

Motion correction results in more uniform

uptake and reduction of motion blurring, especially

in the inferior and anterior regions. The extent of

improvement depends on the amplitude of the heart

motion displacement. A combination of motion

correction with attenuation correction enhanced the

reduction of artifacts. Respiration motion has more

significant effects on the detection of small defects.

In AC-SPECT images, small defects located on

anterior and inferior regions become more visible

due to motion-induced artifacts. AC-SPECT images

with and without motion correction should be

simultaneously inspected to be able to report on

small defects.

This study is limited by the fact that it simulates

only the cranio-caudal displacement of the heart during

respiration. Also, defect detectability in the lateral and

septal wall was not investigated. The respiratory phases

were acquired sequentially and not continuously as

observed in real patients. Further limitations may be due

to phantom assembly (e.g., chest wall expansion, liver

and heart twisting).

Future studies will investigate more realistic condi-

tions of respiratory patterns and cardiac motion

amplitudes as well as defects in other regions of the

LV. Ultimately, such effects shall be studied in future by

investigating real patient cohorts with respiratory-gated

acquisition.

Table 3. Evaluation of physicians’ report (TP, FN)for the 8 small defects

ACstatic

ACnormal

AC1MCnormal

ACdeep

AC1MCdeep

FN 4 0 4 0 4

TP 4 8 4 8 4

AC and MC were applied on SPECT MPI data for phantoms instatic, normal, and deep breathing modes

Journal of Nuclear Cardiology� Polycarpou et al

SPECT myocardial perfusion imaging

Acknowledgments

We gratefully acknowledge the Nuclear Medicine

Department of the Nicosia General Hospital (Andreas

Mormoris, Chariklia Christodoulou, Lambros Lambrou,

Charoula Charalambous) for helping in acquisitions. We

also thank Guy’s and St Thomas Hospitals NHS Foundation

Trust (Steve Turner, Bill Stevens) for optimizing the cardiac

holder. The Project TCEIA/DTCEIA/0311(BIE)/27, with

Host Organization the Frederick Research Center and

Partner Organizations the Cyprus Ministry of Health and the

King’s College London, was co-financed by the European

Regional Development Fund and the Republic of Cyprus

through the Research Promotion Foundation.

Disclosure

The authors declared that they have no conflict of

interest.

References

1. Goetze S, Brown TL, Lavely WC, Zhang Z, Bengel FM. Atten-

uation correction in myocardial perfusion SPECT/CT: Effects of

misregistration and value of reregistration. J Nucl Med

2007;48(7):1090-5.

2. Fricke H, Fricek E, Weise R, Kammeier A, Lindner O, Burchert

W. A method to remove artifacts in attenuation-corrected

myocardial perfusion SPECT introduced by misalignment

between emission scan and CT-derived attenuation maps. J Nucl

Med 2004;45:1619-25.

3. Pitman AG, Kalff V, Van-Every B, Risa B, Barnden LR, Kelly

MJ. Effect of mechanically simulated diaphragmatic respiratory

motion on myocardial SPECT processed with and without atten-

uation correction. J Nucl Med 2002;43:1259-67.

4. Tsui BMW, Segars WP, Lalush DS. Effects of upward creep and

respiratory motion in myocardial SPECT. IEEE Trans Nucl Sci

2000;47:1192-5.

5. Wheat JM, Currie GM. Impact of patient motion on myocardial

perfusion SPECT diagnostic integrity: Part 2. J Nucl Med Technol

2004;32:158-63.

6. Yang YW, Chen JC, He X, Wang SJ, Tsui BM. Evaluation of

respiratory motion effect on defect detection in myocardial per-

fusion SPECT: A simulation study. IEEE Trans Nucl Sci

2009;56:671-6.

7. Segars WP, Tsui BMW. Study of the efficacy of respiratory gating

in myocardial SPECT using the new 4DNCAT phantom. IEEE

Nucl Sci 2002;49:675-9.

8. Kovalski G, Israel O, Keidar Z, Frenkel A, Sachs J, Azhari H.

Correction of heart motion due to respiration in clinical myocar-

dial perfusion SPECT scans using respiratory gating. J Nucl Med

2007;48:630-6.

9. Konik A, Kikut J, Lew R, Johnson K, King MA. Comparison of

methods of acquiring attenuation maps for cardiac SPECT in the

presence of respiratory motion. J Nucl Cardiol 2013;20:1093-107.

10. Ko C, Wu Y, Cheng M, Yen R, Wu W, Tzen K. Data-driven

respiratory motion tracking and compensation in CZT cameras: A

comprehensive analysis of phantom and human images. J Nucl

Cardiol 2015;22:308-18.

11. Livieratos L, Rajappan K, Stegger L, Schafers K, Bailey DL,

Camici PG. Respiratory gating of cardiac PET data in list-mode

acquisition. Eur J Nucl Med Mol Imaging 2006;33:584-8.

12. Chrysanthou-Baustert I, Parpottas Y, Demetriadou O, Christofides

S, Yiannakkaras C, Kaolis D, et al. Diagnostic sensitivity of

SPECT myocardial perfusion imaging using a pumping cardiac

phantom with inserted variable defects. J Nucl Cardiol

2013;20(4):609-15.

13. Cook LB. Rebreathing in the Mapleson A, C, and D breathing

systems with sinusoidal and exponential wavefronts. Anaeshesia

1997;52:1182-94.

14. Rosset A, Spadola L, Ratib O. OsiriX: An open-source software

for navigating in multidimensional DICOM images. J Digit

Imaging 2004;17(3):205-16.

15. McLeish K, Hill DL, Atkinson D, Blackall JM, Razavi R. A study

of the motion and deformation of the heart due to respiration.

IEEE Trans Med Imaging 2002;21(9):1142-50.

16. Hesse B, Kristina-Tagil A, Cuocolo C, Anagnostopoulos M,

Bardies J, Bax F, et al. EANM/ESC procedural guidelines for

myocardial perfusion imaging in nuclear cardiology. Eur J Nucl

Med Mol Imaging 2005;32(7):855-97.

17. Buerger C, Schaeffter T, King AP. Hierarchical adaptive local

affine registration for fast and robust respiratory motion estima-

tion. Med Image Anal 2011;15:551-64.

18. Axel M, Zikic D, Botnar R, Bundschuh R, Howe W, Ziegler S,

et al. Dual cardiac-respiratory gated PET: Implementation and

results from a feasibility study. Eur J Nucl Med Mol Imaging

2007;34(9):1447-54.

19. Masood Y, Liu Y-H, DePuey G, Taillefer R, Araujo LI, Allen S,

et al. Clinical validation of SPECT attenuation correction using x-

ray computed tomography-derived attenuation maps: Multicenter

clinical trial with angiographic correlation. J Nucl Cardiol

2005;12:676-86.

Polycarpou et al Journal of Nuclear Cardiology�SPECT myocardial perfusion imaging