How to use cardiac IQ•SPECT routinely? An overview of tips and tricks from practical experience to the literature

European Journal of Nuclear Medicine and Molecular Imaging, Dec 2015

Eric Gremillet, Denis Agostini

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

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 How to use cardiac IQ SPECT routinely? An overview of tips and tricks from practical experience to the literature Eric Gremillet 0 1 2 Denis Agostini 0 1 2 0 Department of Nuclear Medicine, University Hospital of Caen and Normandie Université , EA4650 Caen , France 1 Centre d'Imagerie Nucléaire, Hôpital Privé de la Loire , Saint-Etienne , France 2 Eric Gremillet Introduction The main part of the IQ SPECT tomographic system developed by Siemens dedicated to cardiac imaging is a multifocal collimator called SMARTZOOM. The system can be installed on Symbia cameras from the same manufacturer. Besides a pair of SMARTZOOM collimators, it includes cardiocentric acquisition software and reconstruction software that takes into account this particular geometry of detection in an OSEM algorithm and can correct for scatter diffusion. Although presented as optional by the manufacturer, attenuation correction (AC) using a CT scan (with 2 to 16 detectors, depending on the exact model of Symbia camera) is an additional and e s s e n t i a l f e a t u r e . T h e m u l t i f o c a l g e o m e t r y o f SMARTZOOM enables magnification of the centre of the field of view where the heart lies by a factor of up to four while avoiding truncation of lateral parts of the patient’s torso. * Concerning the multifocal acquisition geometry, the manufacturer calls the region of the field of view where the magnification is fourfold the Bsweet spot^. The technologist has to centre the heart as precisely as possible within the sweet spot and the myocardial contrast may often be less than optimal, and thus the effects of mispositioning on the reconstructed image is a concern. This question was addressed by Rajaram et al. [ 1 ] in a phantom study. They found that final images were significantly affected by different positioning (centring on the base vs. the apex) when they were reconstructed without AC (NAC), while no difference was observed with AC. A recent phantom study by Caobelli et al. [ 2 ] confirmed these results. The effects of mispositioning by up to 2.5 cm were considered unlikely to affect clinical results when AC was applied. The reason for this AC dependency can be understood from careful consideration of the ray paths through the patient’s body. For rays originating from the heart base, the path length is longer with a SMARTZOOM collimator than with a parallel collimator, while there is no difference for rays originating from the apex. This phenomenon, which is purely geometrical, can vary with the heart positioning around the sweet spot region. Thus attenuation artefacts are not only more marked than with parallel collimation (as discussed below) but also unpredictable and position-dependent, making AC mandatory. Users of a Symbia camera not equipped with a CT scanner who are considering installing an IQ SPECT system should be aware of this limitation. Regarding detection characteristics, IQ SPECT has been compared with parallel collimation and with two dedicated CZT cameras in phantom and human imaging [ 3 ]. Four parameters were analysed with the systems under study: sensitivity, resolution, contrast/noise ratio, and image sharpness. To summarize, CZT cameras are better than Anger cameras for all parameters, and IQ SPECT is not different from parallel collimation except for its sensitivity which is four times better, as expected. Attenuation correction NAC images obtained with IQ SPECT often exhibit a hot apex and poor contrast in basal segments due to attenuation, while AC images usually show a more homogeneous pixel intensity throughout the left ventricle wall. We have previously discussed a likely explanation for this phenomenon: the basal parts of the ventricle are more prone to attenuation artefacts than with conventional parallel collimation. But apex thinning is often observed in AC images. In some cases the apex even looks quite normal in NAC images but unusually thin in AC images. In addition to correcting for the particular attenuation artefacts of IQ SPECT, AC allows correction of common attenuation artefacts such as low pixel counts in the inferior wall. AC also makes it possible to obtain satisfying images in very obese patients or in patients with the left arm along the body. The additional radiation dose from CT may be a concern. On the one hand, our physicists have found that the CT voltage can be reduced to 110 kV (in patients up to 130 kg) without loss of AC accuracy (unpublished data from Hopital prive de la Loire), and on the other hand, we have demonstrated that the same CT acquisition can be used for both stress and rest AC if the CT and rest SPECT images are carefully manually registered. However, it is reasonable to obtain a separate CT scan at rest in patients in whom similar repositioning is very uncertain (for example, very large breast, left arm kept along the body, marked kyphosis). Volumes and LVEF assessment Using the Amsterdam gated (AGATE) dynamic cardiac phantom (Vanderwilt techniques, Boxtel, The Netherlands), Bailliez et al. [ 4 ] used CZT cameras and IQ SPECT to determine volumes and left ventricular ejection fraction (LVEF). Eighteen acquisitions were processed on each CZT and Anger (IQ SPECT) camera. New CZT cameras yielded different results for global left ventricular function compared to Anger camera with cardiofocal collimators. Left ventricular volumes were higher with the DNM 530c than with D-SPECT and IQ SPECT, leading to decreased LVEF in normal subjects. These differences should be taken into account in clinical practice and indicate the need for the collection of a specific normal database. Clinical results As an expected consequence of the fourfold increase in sensitivity, the first published clinical work comparing IQ SPECT and conventional parallel collimation showed that the acquisition duration could be divided by approximately four with similar image quality and diagnostic accuracy. Other workers have confirmed this finding in both phantom studies [ 5 ] and clinical studies [ 6 ]. The phantom study also confirmed that either the activity or the acquisition duration can be reduced. While all the previously mentioned studies used 99mTc tracers, the technique has also been used and validated with 201Tl [ 7, 8 ]. In both studies with 201Tl, concordance between IQ SPECT CT and conventional parallel collimation was better when AC was used. The second study also showed that inferior wall attenuation artefacts often observed with 201Tl and conventional parallel collimation are improved by AC. Thus overall, it appears that AC IQ SPECT gives clinical results similar to those with conventional parallel collimation whatever the tracer, while allowing the Binjected dose × image duration^ product to be divided by 4. Despite the fact that contrast/noise ratio and image sharpness are not improved with IQ SPECT, image quality may be improved over conventional parallel collimation. Indeed, since acquisition duration is greatly reduced, the likelihood of motion (whether of the heart within the patient, e.g. upward creep, or of the patient) is also reduced. Thus motion artefact is less likely to occur. It has been known for a long time that such artefacts can be very subtle, are unpredictable and can mimic true abnormalities. ECG gating IQ SPECT with ECG gating often produces unexpected results for LVEF. Rigorous results on this particular point have been found in only one study in 80 patients [ 9 ] in which functional parameters obtained with IQ SPECT were compared with those from parallel collimation. LVEF was significantly lower with IQ SPECT: −8 % on stress images, −9 % on rest images. Since this was related to slightly higher left ventricular volumes overall, the authors hypothesized that scatter correction which is included in IQ SPECT reconstruction software could indirectly have been responsible by improving resolution, which can result in slightly higher cavity volumes both at end diastole and end systole. However, we have previously stated that resolution is actually identical with IQ SPECT and parallel collimation, which is not in favour of this hypothesis. The findings of the previous study [ 9 ] were not dependent upon the contouring software used (4 DM SPECT) because similar results have been observed by many others, including our group, using concurrent software (QGS), although no data have been published on this topic. Protocols It is possible to make the most of the increased sensitivity of IQ SPECT by shortening stress–rest protocols. The basic ideas behind such protocols are: 1. Stress injected activity can be greatly reduced while maintaining good image quality. 2. It is thus possible to greatly increase the rest injected activity in a given patient while keeping the total daily activity within recommended limits [ 10 ]. 3. The ratio between rest and stress injected activities is thus increased. 4. Finally, the delay between injections can be reduced while keeping contamination of the rest image by residual stress counts at an acceptable level. By combining the results of several substudies it is possible to design a short 1-day protocol (stress then, if necessary, rest). For example, for a 70-kg patient, the delay between stress and rest injections is reduced to 1 h and the acquisition duration is 9 min under stress (to which must be added around 5 min for the CT acquisition) and 3 min at rest (without CT acquisition). The reconstruction OSEM parameters routinely used are 10 iterations and 3 subsets for summed nongated images and 12 iterations and 1 subset for gated images, with a 10-mm Gaussian filter and scatter correction for both. Takamura et al. [ 11 ] investigated whether prone myocardial perfusion SPECT with 201Tl acquired with IQ-SPECT could correct for soft-tissue attenuation in 39 patients. They concluded that attenuation was better and defect decisions were similar with prone IQ-SPECT images than with supine IQSPECT images in the anterior, anteroseptal, lateral and inferior areas. Tips and tricks Although the interpretation of IQ SPECT images is already well understood by experienced IQ SPECT users and experts in nuclear cardiology, the following tips on interpretaion may be helpful to future users. & Prefer prone positioning for acquisition whenever possible: – – Inferior wall attenuation artefacts are less marked on NAC images and NAC images must be considered in the interpretation. There is less motion in the prone position than in the supine position (heart within torso and/or patient on table). Quality control: check the acquisition data for quality of ECG gating, centring and motion. Since the acquisition duration is short, do not hesitate to repeat if necessary (the CT part does not need to be repeated). CT and SPECT image registration: – – – – Check the CT images for artefacts, in particular for the presence of a notch on the left margin of the heart, a finding that is not uncommon with single-slice or two-slice CT scans. This can produce marked artefacts in AC images, mimicking true hypoperfusion. When the CT scan is acquired at the same time, generally no user action is required, or there is a shift of a few millimetres. When using a separately acquired CT scan, manual registration may involve a shift of a few centimetres in all three directions. Adjust the colour scale such that the left myocardium is well seen, look only at the coronal and transverse plane displays (the sagittal image can be misleading due to variable overlap of the right and left ventricles), use the arrow keys (more precise than the mouse), and make wide and frequent use of the blend slider to check registration. If an abnormality is seen only on the AC images while the NAC images look normal, then an AC artefact is very likely (e.g. due to misregistration of CT and SPECT images since alignment cannot always be perfect, or due to a notch artefact). If a nonreversible apical thinning pattern is seen (a frequent finding), look carefully at the beating slices since evidence of akinesis, or at least hypokinesis, of the apex can lead to the (rare) conclusion that a small nontransmural or transmural apical scar is present. Similarly, if the apical thinning pattern is partially reversible, an artefact is likely. More generally, as a rule of thumb, if abnormalities are seen on AC images, always cross-check the AC images with the NAC images. If the abnormalities are real at least a faint similar trend should be observed. The aim of AC is to correct for artefactual patterns due to attenuation problems, whether general such as diaphragmatic or SMARTZOOM-specific as discussed above. Since gated images are not corrected for attenuation, they sometimes, and perhaps often, exhibit a hot apex and a faint base. Moreover, this pattern may be more or less marked depending on the time bin. Thus it is & Perfusion nongated images: & Functional gated images: – more reliable and easier to visually assess wall motion rather than wall thickening. We use the three-plane display called Slice in QGS, setting the colour to grey and the gamma to a high value in the range 1.5 to 1.7. To limit the problem of an erroneously small LVEF, avoid masking out the heart on gated slices or, if masking is necessary, avoid masking too close to the left ventricle. Conclusion Compared with conventional parallel collimation, IQ SPECT shows similar global image quality, including resolution, but a fourfold higher sensitivity which allows the use of short 1-day protocols. Since attenuation artefacts are more problematic than with conventional parallel collimation, we strongly advise the use of CT attenuation correction. Taking into account these precautions and the interpretation tips, IQ SPECT may be a good cardiac imaging modality that would increase the throughput of patients with coronary artery disease. 1. Rajaram R , Bhattacharya M , Ding X , Malmin R , Rempel TD , Vija AH , et al. Tomographic performance characteristics of the IQ·SPECT system . IEEE Nuclear Science Symposium Conference Record; 2011 . p. 2451 - 6 . doi: 10 .1109/NSSMIC. 2011 . 6152666 2. 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 ( 1 ): 57 - 65 . 3. Imbert L , Poussier S , Franken PR , Songy B , Verger A , Morel O , et al. Compared performance of high-sensitivity cameras dedicated to myocardial perfusion SPECT: a comprehensive analysis of phantom and human images . J Nucl Med . 2012 ; 53 : 1897 - 903 . 4. Bailliez A , Lairez O , Merlin C , Piriou N , Legallois D , Blaire T et al. Global and regional left ventricular function assessment using two different semiconductor cadmium zinc telluride (CZT) cameras compared to conventional gamma camera with cardiofocal collimators: dynamic cardiac phantom study and clinical validation . J Nucl Med 2015 (abstract SNM) . 5. Caobelli F , Ren Kaiser S , 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 . 6. Caobelli F , Pizzocaro C , Paghera B , Guerra UP . Evaluation of patients with coronary artery disease: IQ-SPECT protocol in myocardial perfusion imaging - preliminary results . Nuklearmedizin . 2013 ; 582 : 178 - 85 . 7. Horiguchi Y , Ueda T , Shiomori T , Kanna M , Matsuhita 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 . 8. 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 . 9. Pirich C , Keinrath P , Barth G , Rendl G , Rettenbacher L , Rodrigues M. Diagnostic accuracy and functional parameters of myocardial perfusion scintigraphy using accelerated cardiac acquisition with IQ SPECT technique in comparison to conventional imaging . Q J Nucl Med Mol Imaging . 2014 (Epub ahead of print). 10. Verberne HJ , Acampa W , Anagnostopoulos C , Ballinger J , Bengel F , De Bondt P , et al. EANM procedural guidelines for radionuclide myocardial perfusion imaging with SPECT and SPECT/CT: 2015 revision . Eur J Nucl Med Mol Imaging . 2015 ; 42 ( 12 ): 1929 - 40 . 11. 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 tomographyderived attenuation correction . Ann Nucl Med . 2015 ; 29 ( 10 ): 890 - 6 .

This is a preview of a remote PDF:

Eric Gremillet, Denis Agostini. How to use cardiac IQ•SPECT routinely? An overview of tips and tricks from practical experience to the literature, European Journal of Nuclear Medicine and Molecular Imaging, 2016, 707-710, DOI: 10.1007/s00259-015-3269-1