New techniques, distinctive population, unique normal databases
New techniques, distinctive population, unique normal databases
James R. Galt 0 2
Ph.D 0 1 2
0 Reprint requests: James R. Galt, Ph.D., Department of Radiology and Imaging Sciences, Emory University Hospital , 1364 Clifton Road, Atlanta, Georgia 30322 , USA
1 Department of Radiology and Imaging Sciences, Emory University School of Medicine , Atlanta , Georgia
2 Galt New techniques , distinctive population
No charge is more important to nuclear cardiology
than that of efficiently providing accurate results at low
radiation doses. Keeping the radiation dose low requires
that we do a better job of detecting photons emitted from
the myocardium and do a better job of utilizing each
photon we detect. This quest has drastically increased
the diversity of reconstruction techniques and
instrumentation for SPECT myocardial perfusion imaging.
Each reconstruction technique produces images with its
own unique texture. Each novel device introduces new
considerations for the technologist acquiring the scan
and new considerations for the clinician interpreting the
images including unique patterns of normal and unique
artifacts. When quantitative analysis is performed,
normal databases developed with standard equipment
and techniques may no longer be appropriate.1,2 The
study that inspired this editorial details the construction
of 201Tl normal databases for one of one such system,
IQ SPECT, in a distinctive population (Japan).3
Several methods have been developed to shorten
acquisition time, to allow a reduction in injected
radiopharmaceutical dose or a trade-off between the two.
Originally developed to provide short scan times,
advanced reconstruction methods seek to do a better job
of producing SPECT images with the counts acquired by
the imaging system than filtered backprojection or
standard iterative reconstruction methods. As the
improvements work with conventional SPECT systems,
they can be thought of as providing an effective increase
in photon sensitivity. These algorithms model the
acquisition process including the geometric response
of the collimator (sometimes called resolution recovery)
and use unique noise reduction algorithms to produce
high-quality images with a substantial reduction in
Other systems have increased photon sensitivity
through the use of high-sensitivity collimation and
multiple detectors to perform a cardio-centric
acquisition. Regardless of the collimation used, each system
also takes advantage of advanced SPECT reconstruction
algorithms to further improve the sensitivity of the
One such system uses three pixelated cesium iodide
scintillation detectors and fan-beam collimation to
increase sensitivity of counts collected from the heart
(Cardius X-ACT, Digirad, Poway, CA). In this system,
the patient sits upright in a chair that rotates. During
acquisition, the heart of a properly set-up patient
remains in the center of each of the fan-beam
collimators, maximizing counts from the heart.
Two other systems utilize direct-conversion
solidstate detectors made of Cadmium Zinc Telluride (CZT)
but very different collimation schemes. The first CZT
system uses nine rectangular detectors with very
highsensitivity parallel-hole collimation that sweep across
the heart of the patient (D-SPECT Spectrum Dynamics,
Caesarea, Israel). Counts from the heart are maximized
by confining the sweep of each detector to the region of
the heart. The second CZT system uses 19 square CZT
detectors each equipped with a single pinhole collimator
(GE Healthcare, Haifa, Israel). The center line of each
detector passes through the same point in space. With
this system the detectors and the patient remain
stationary during acquisition and the heart of a properly
positioned patient is centered on the point where the
center lines of all the pinhole collimators meet.
The system used in the paper by Okuda achieves
high sensitivity using a unique collimator design on a
standard, large field of view, dual detector SPECT or
SPECT/CT system (IQ SPECT, Siemens, Erlangen,
Germany).3 The confocal collimators used in IQ SPECT
have a central area with converging collimation
designed to focus on the heart and transition to
parallel-hole collimation around the periphery of the camera.
This design allows increased sensitivity over the heart
where it is most needed but avoids truncation of the
body.7 It is important that the heart be properly
positioned in the region of highest magnification of
both collimators throughout acquisition. While slight
mispositioning may be tolerable, poor patient set-up
may be less forgiving than standard parallel-hole
collimation.8,9 It has also been noted that attenuation artifacts
may differ from those routinely recognized with
standard parallel-hole collimation and that attenuation
corrected images may be preferred.9
The use of IQ SPECT to reduce imaging time and/
or radiopharmaceutical dose has been demonstrated in
phantoms and in patients.7,10 In one study based on
patient images, Lyon et al compared attenuation
corrected stress SPECT to IQ SPECT using a dose of
9251100 MBq (25-30 mCi) Tc-99m sestamibi. Several
different count levels were simulated for IQ SPECT, and
evaluated using system and count level specific normal
files. The study concluded that IQ SPECT could be used
to reduce both the dose and the time by half compared to
conventional SPECT. Thus, the standard dose could be
reduced to below 550 MBq (15 mCi) and the imaging
time reduced from 13 minutes (standard SPECT) to 7
minutes (IQ SPECT).10
While the three high-sensitivity hardware designs
discussed above are dedicated cardiac cameras, only
when the confocal collimator is mounted is the system
restricted to nuclear cardiology. When equipped with
other collimators, the system is a general purpose
SPECT or SPECT/CT, a factor that may appeal to
clinics that also perform general nuclear medicine
Nuclear cardiology professionals, both
technologists and physicians, can expect a significant learning
curve when switching to high-sensitivity SPECT. Patient
setup is not only different for the technologist but can be
much less forgiving of error when the cameras and
collimation need to focus on the heart. Physicians may
find that images may have unfamiliar texture, and clues
to separating artifacts from normality, such as normal
attenuation patterns, may also differ.
Properly implemented, quantitative analysis software
that compares each scan with normal databases can not
only serve as tool to help with diagnoses but may also
help a physician become comfortable with an unfamiliar
system. This only holds true, however, if normal
databases for your SPECT system, reconstruction
software, radiopharmaceutical, and patient demographics are
available.1,11–15 When quantitative comparison to normal
databases is applied, normal databases need to match the
imaging system, the radiopharmaceutical, the
reconstruction techniques, and the patient population.
The Japanese Society of Nuclear Medicine working
group (JSNM-WG) has worked to standardize
techniques and has developed databases for the Japanese
population using conventional SPECT systems.16–18 A
comparison of the Japanese normal databases to a US
database generated with conventional SPECT found
significant differences. The authors concluded that not
only was an acquisition specific database essential but it
was essential to use population-specific databases as
well. The authors also speculated that the differences
might have been due to differences in body habitus,
rather than ethnicity. BMIs reported in that paper were
22 ± 3 and 24 ± 3 for the Japanese database and 27 ± 5
for the US database.19 Extrapolation of from Okuda
Table one 3 indicates the average BMI in the Japanese
IQ SPECT database would be similar or smaller than
those of the Japanese standard database.
201TL IQ SPECT IN JAPAN
The use of IQ SPECT with 201Tl in a Japanese
population has been reported previously. Horiguchi et al
showed that, compared to standard acquisition of
20 minutes, that image quality and semi-quantitative
analysis of IQ SPECT with an acquisition of 8 minutes
yielded comparable results, especially when CTAC was
used.20 Takamura et al investigated the use of prone
imaging with IQ SPECT, finding that prone imaging
provided similar benefits for avoiding artifacts as
CTAC. Acquisition times were 8 minutes.21 Matsuo
et al compared IQ SPECT acquisitions of 6 minutes
with a standard acquisition of 20 minutes finding that
the images were of equivalent quality. They noted that
CTAC improved inferior artifacts with some apparent
decrease in anterior or apical anterior segments.22
Standard doses set by the Japanese working group
for Tl-201 are 74-111 MBq (2-3 mCi), which is used in
about half of the SPECT MPI studies as of 2015 due to
its high extraction fraction, defect contrast, image
quality, and the use of a single administration for both
stress and rest.18 Each of the studies at total of 111 MBq
(3 mCi) 201Tl was used. Horiguchi and Takamura
injected the full dose at stress and performed delayed
imaging at 3 hours. The delayed images were used in
their investigations. Matsuo injected 74 MBq (2 mCi) at
stress. Resting images were acquired 2–3 hours after an
additional injection of 37 mBq (1 mCi).
Use of 201Tl at these doses makes it difficult to meet
the goal of an effective radiation dose of B9 mSv given
in the ASNC Information Statement on
Recommendations for Reducing Radiation Exposure in Myocardial
Perfusion Imaging.23 Using an estimate of 4.4 mSv per
mCi (37 mBq)24 74 MBq yields 8.8 mSv. Increasing the
dose to 111 mBq yields 13.2 mSv. Adding a CT dose of
up to 2.5 mSv for attenuation correction10 pushes the
total dose well over the 9 mSv limit.
Assuming that use of 201Tl myocardial SPECT will
continue and that IQ SPECT is used at a significant
number of imaging centers, the uniqueness of the
collimation, the radiopharmaceutical used, and the
patient demographics made the development of new
normal databases necessary. Okuda et al set out to
develop these databases for various protocols: rest and
stress acquisitions for supine, prone, and ACSC. This
work was done with retrospective data and without
rigorous standardization.3 Ideally, this would be done
prospectively with standard acquisition protocols but
sometimes you have to work with the data you have.
The significant differences between the IQ SPECT
normal databases developed for the different acquisition
protocols illustrate the need for the separate databases.
The normal databases developed for confocal
collimation and the unique reconstruction differ significantly
different from those developed for conventional SPECT
(as well as the differences between the male and female
normal files—or the lack of difference with ACSC).
Others, including this editorialist, have found that when
attenuation correction is applied, a combined normal
database may be used for male and female patients.25,26
The use of new high-sensitivity SPECT comes at
the cost of images that are different from those that
familiar to the clinician. At the same time, application of
quantitative software to these images requires the
development of new normal databases. In the US, most
users are accustomed to receiving appropriate databases
with the purchase of quantitative software. Those
databases may not be available with the introduction
of new imaging equipment, new radiopharmaceuticals,
or as old radiopharmaceuticals (201Tl) fall out-of-favor.
James R. Galt has no disclosures related to this editorial.
In areas of the world with a population that differs in
body habitus from that developed for the US, the need
may be even greater. The bottom line is that the benefits
of quantitative software come at the cost of normal
database development for different imaging systems,
reconstruction techniques, radiopharmaceuticals, and
1. Zoccarato O , Marcassa C , Lizio D , Leva L , Lucignani G , Savi A , et al. Differences in polar-map patterns using the novel technologies for myocardial perfusion imaging . J Nucl Cardiol 2016 . doi:10.1007/s12350- 016 - 0500 -9.
2. Slomka PJ , Rubeaux M , Germano G . Quantification with normal limits: New cameras and low-dose imaging . J Nucl Cardiol 2016 . doi:10.1007/s12350- 016 - 0563 -7.
3. Okuda K , Nakajima K , Matsuo S , Kondo C , Sarai M , Horiguchi Y , et al. Creation and characterization of normal myocardial perfusion imaging databases using the IQ SPECT system . J Nucl Cardiol 2017 . doi:10.1007/s12350- 016 - 0770 -2.
4. Gordon DePuey E 3rd. Advances in cardiac processing software . Semin Nucl Med 2014 ; 44 : 252 - 73 . doi: 10.1053/j.semnuclmed. 2014 .04.001.
5. Piccinelli M , Garcia EV . Advances in single-photon emission computed tomography hardware and software . Cardiol Clin 2016 ; 34 : 1 - 11 . doi:10.1016/j.ccl. 2015 .06.001.
6. Slomka PJ , Berman DS , Germano G. New cardiac cameras: single-photon emission CT and PET . Semin Nucl Med 2014 ; 44 : 232 - 51 .
7. Caobelli F , Kaiser SR , Thackeray JT , Bengel FM , Chieregato M , Soffientini A , et al. IQ SPECT allows a significant reduction in administered dose and acquisition time for myocardial perfusion imaging: evidence from a phantom study . J Nucl Med 2014 ; 55 : 2064 - 70 .
8. Caobelli F , Ren Kaiser S , Thackeray JT , Bengel FM , Chieregato M , Soffientini A , et al. The importance of a correct positioning of the heart using IQ-SPECT system with multifocal collimators in myocardial perfusion imaging: A phantom study . J Nucl Cardiol 2015 ; 22 : 57 - 65 .
9. Gremillet E , Agostini D. How to use cardiac IQ SPECT routinely? An overview of tips and tricks from practical experience to the literature . Eur J Nucl Med Mol Imaging 2016 ; 43 : 707 - 10 .
10. Lyon MC , Foster C , Ding X , Dorbala S , Spence D , Bhattacharya M , et al. Dose reduction in half-time myocardial perfusion SPECT-CT with multifocal collimation . J Nucl Cardiol 2016 ; 23 : 657 - 67 .
11. Slomka PJ , Rubeaux M , Germano G . Quantification with normal limits: New cameras and low-dose imaging . J Nucl Cardiol 2016 . doi:10.1007/s12350- 016 - 0563 -7.
12. Miao TL , Kansal V , Glenn Wells R , Ali I , Ruddy TD , Chow BJ . Adopting new gamma cameras and reconstruction algorithms: Do we need to re-establish normal reference values ? J Nucl Cardiol 2016 ; 23 : 807 - 17 .
13. DePuey EG . Sources of variability of gated myocardial perfusion SPECT quantitative parameters . J Nucl Cardiol 2016 ; 23 : 818 - 23 .
14. Alexiou S , Georgoulias P , Angelidis G , Valotassiou V , Tsougos I , Psimadas D , et al. Myocardial perfusion and left ventricular quantitative parameters obtained using gated myocardial SPECT: Comparison of three software packages . J Nucl Cardiol 2016 . doi: 10.1007/s12350- 016 - 0730 -x.
15. Germano G . Quantitative measurements of myocardial perfusion and function from SPECT (and PET) studies depend on the method used to perform those measurements . J Nucl Cardiol 2016 . doi:10.1007/s12350- 016 - 0757 -z.
16. Nakajima K , Matsuo S , Kawano M , Matsumoto N , Hashimoto J , Yoshinaga K , et al. The validity of multi-center common normal database for identifying myocardial ischemia: Japanese Society of Nuclear Medicine working group database . Ann Nucl Med 2010 ; 24 : 99 - 105 .
17. Nakajima K. Normal values for nuclear cardiology: Japanese databases for myocardial perfusion, fatty acid and sympathetic imaging and left ventricular function . Ann Nucl Med 2010 ; 24 : 125 - 35 .
18. Nakajima K , Matsumoto N , Kasai T , Matsuo S , Kiso K , Okuda K. Normal values and standardization of parameters in nuclear cardiology: Japanese Society of Nuclear Medicine working group database . Ann Nucl Med 2016 ; 30 : 188 - 99 .
19. Nakajima K , Okuda K , Kawano M , Matsuo S , Slomka P , Germano G , et al. The importance of population-specific normal database for quantification of myocardial ischemia: Comparison between Japanese 360 and 180-degree databases and a US database . J Nucl Cardiol 2009 ; 16 : 422 - 30 .
20. Horiguchi Y , Ueda T , Shiomori T , Kanna M , Matsushita H , Kawaminami T , et al. Validation of a short-scan-time imaging protocol for thallium-201 myocardial SPECT with a multifocal collimator . Ann Nucl Med 2014 ; 28 : 707 - 15 .
21. Takamura T , Horiguchi Y , Kanna M , Matsushita H , Sudo Y , Kikuchi S , et al. Validation of prone myocardial perfusion SPECT with a variable-focus collimator versus supine myocardial perfusion SPECT with or without computed tomography-derived attenuation correction . Ann Nucl Med 2015 ; 29 : 890 - 6 .
22. Matsuo S , Nakajima K , Onoguchi M , Wakabayash H , Okuda K , Kinuya S. Nuclear myocardial perfusion imaging using thallium201 with a novel multifocal collimator SPECT/CT: IQ-SPECT versus conventional protocols in normal subjects . Ann Nucl Med 2015 ; 29 : 452 - 9 .
23. Cerqueira MD , Allman KC , Ficaro EP , Hansen CL , Nichols KJ , Thompson RC , et al. Recommendations for reducing radiation exposure in myocardial perfusion imaging . J Nucl Cardiol 2010 ; 17 : 709 - 18 .
24. Henzlova MJ , Duvall WL , Einstein AJ , Travin MI , Verberne HJ . ASNC imaging guidelines for SPECT nuclear cardiology procedures: Stress, protocols, and tracers . J Nucl Cardiol 2016 ; 23 ( 3 ): 606 - 39 .
25. Grossman GB , Garcia EV , Bateman TM , Heller GV , Johnson LL , Folks RD , et al. Quantitative Tc-99m sestamibi attenuation-corrected SPECT: Development and multicenter trial validation of myocardial perfusion stress gender-independent normal database in an obese population . J Nucl Cardiol 2004 ; 11 : 263 - 72 .
26. Esteves FP , Galt JR , Folks RD , Verdes L , Garcia EV . Diagnostic performance of low-dose rest/stress Tc-99m tetrofosmin myocardial perfusion SPECT using the 530c CZT camera: quantitative vs visual analysis . J Nucl Cardiol 2014 ; 21 : 158 - 65 .