Taking the perfect nuclear image: Quality control, acquisition, and processing techniques for cardiac SPECT, PET, and hybrid imaging

Abstract

Nuclear Cardiology for the past 40 years has distinguished itself in its ability to non-invasively assess regional myocardial blood flow and identify obstructive coronary disease. This has led to advances in managing the diagnosis, risk stratification, and prognostic assessment of cardiac patients. These advances have all been predicated on the collection of high quality nuclear image data. National and international professional societies have established guidelines for nuclear laboratories to maintain high quality nuclear cardiology services. In addition, laboratory accreditation has further advanced the goal of the establishing high quality standards for the provision of nuclear cardiology services. This article summarizes the principles of nuclear cardiology single photon emission computed tomography (SPECT) and positron emission tomography (PET) imaging and techniques for maintaining quality: from the calibration of imaging equipment to post processing techniques. It also will explore the quality considerations of newer technologies such as cadmium zinc telleride (CZT)-based SPECT systems and absolute blood flow measurement techniques using PET.

Introduction

Performance of high quality nuclear cardiac imaging is demanding. Equipment must be up to date and calibrated. Lab personnel need to be trained on pre-imaging quality-control steps, the imaging protocol, data acquisition and processing. In addition, these steps may need to be optimized for each patient, while avoiding short-cuts and improvisation. Finally, interpretation of the images must include quality assessment of instrumentation performance and adherence to protocol in order to differentiate artifact from disease.^1,2

Perhaps the most significant, intrinsic limitation to nuclear cardiology is the small amount of radioactive tracer that may be injected. The combination of low count data and long acquisition times that are customary in nuclear cardiology are a direct source of poor resolution, motion artifacts, and noise texture when compared to other modalities. In addition, U.S. and international standards for acceptable radiation dosage continue to put downward pressure on what is considered acceptable tracer dosages.3., 4, 5.

To put in perspective the challenge of nuclear imaging, consider that a routine nuclear cardiac perfusion study requires an acquisition time 200,000 times longer than conventional photography, while acquiring only a fraction of the counts per pixel! These long acquisition times make nuclear imaging particularly susceptible to patient motion. Motion artifacts can result from the external patient movement or by changes in the position of the patient’s organs: such as diaphragm relaxation or cardiac motion diaphragm positioning. In a study of 48 patients, 17% of studies had significant motion,6, 7, 8 appearing as everything from a small focal defect to a more significant “hurricane” artifact.⁹ Motion correction programs are ubiquitous in nuclear cardiology, but often times fail to correct motion-related artifacts and can even introduce new artifacts.10, 11, 12

Photon attenuation and scatter is another major source of imaging artifacts and is often cited as the reason for low specificity when compared to coronary angiography.13, 14, 15 Because of this, nuclear laboratories have explored the use of attenuation and scatter correction in cardiac SPECT to improve interpretive accuracy. Studies have reported significant improvements in normalcy and reduction in false-positive rates when attenuation correction is applied.16, 17, 18, 19, 20 The benefit of attenuation and scatter correction may be more profound in targeted populations such as women and the obese.²¹ It may also enable other applications such as stress only imaging.^22,23 Because of this, the American Society of Nuclear Cardiology and Society of Nuclear Medicine have issued a joint statement on attenuation correction recommending it when possible.²⁴ Despite these advantages and recommendations, attenuation and scatter correction for SPECT has not become widely adopted in the clinical setting.²⁵ In contrast to SPECT, nearly all commercially available PET imaging systems are equipped with attenuation correction apparatus, and attenuation and scatter correction are applied in nearly all cases. The routine application of attenuation correction in PET has been cited as a major reason for its higher specificity compared to SPECT.26, 27, 28, 29

Detector blur and partial volume effects also can have an impact on the images, reducing sensitivity to defects and resulting in inaccurate measurements of quantities, such as LV volumes.^30,31 It has been demonstrated that partial volume artifacts can be reduced when a reconstruction zoom is employed³⁰ and when smaller pixel sizes are used.³¹ In cardiac PET, partial volume effects can create hot spots in the papillary muscles and a fixed perfusion defect at the apex, presumably due to apical thinning along with dynamic motion.³² Though the impact of PET partial volume effects are less significant for visual assessment, partial volume effects can introduce significant errors in absolute quantitative measurements, such as myocardial flow and flow reserve.^33,34

High quality nuclear cardiology can be obtained by understanding the sources of artifact, implementing strategies that minimize their impact and judiciously correcting for those artifacts when necessary.