SMARTZOOM™ HRX White Paper

White paper SMARTZOOM HRX Applications and benefits in SPECT and SPECT/CT brain imaging Brandon Clinton Jones, PhD Francesc Massanés, PhD A. Hans Vija, PhD siemens-healthineers.com/symbia-prospecta SIEMENS Healthineers Table of contents Overview 3 Clinical applications 4 Neurological applications 4 Other clinical applications 5 Brain scan positioning 6 Performance and validation 7 Resolution benchmarking 7 Validation data for neurological applications 8 Collimator design and image reconstruction 10 Design 10 Automated quality control for all manufactured collimators 11 Organ-centric acquisition 11 Image reconstruction 11 SMARTZOOM HRX considerations 13 About the lead author 13 References 14 2 White paper · SMARTZOOM HRX Overview: SMARTZOOM HRX SMARTZOOM™ HRX with high resolution and extended magnification volume was introduced in 2022 as an addition to the SMARTZOOM line of multifocal collimators. The SMARTZOOM product line is a class of multifocal collimators that adapts general- purpose SPECT or SPECT/CT scanners to organ-focused protocols with improved imaging sensitivity.1-6 Multifocal collimators “magnify” the organ of interest by casting the projection image of the organ from a small region of interest (ROI) onto a larger portion of the detector as compared to non-magnifying collimators, while still avoiding image truncation with near-parallel collimation along the periphery. In this way, multifocal collimators increase the sensitivity for the organ of interest, at the cost of diminished sensitivity near the edge of the field of view (FOV), which is commonly of little clinical interest. The depth-dependent focusing region, which produces a magnification >1, is colloqui- ally referred to as the “sweet spot.” Multifocal collimators have several key advantages over conventional fanbeam colli- mators. Firstly, fanbeam collimators suffer from image truncation near the image boundary, which reduces image quality and limits their use outside of neurological applications. Conversely, multifocal collimators avoid image truncation using near- parallel collimation on the image periphery and are thus able to tackle numerous clinical applications throughout the body. Secondly, multifocal collimators inherently allow for “magnification” along both detector dimensions (transaxial and longitudinal), which are multiplied together to produce a larger total magnification. Since fanbeam collimators are restricted to magnification only along the transaxial dimension, the excessive tilt in the collimator holes needed to achieve the same effec- tive magnification complicates the image reconstruction and renders them ill-suited for iterative reconstruction algorithms with resolution recovery. Multifocal collimators, on the other hand, are well-suited for efficient iterative reconstruction algorithms with resolution recovery and achieve greater performance with less noise buildup, which is a prerequisite for multi-modal or quantitative SPECT imaging. The first SMARTZOOM collimator, powered by Siemens Healthineers IQ•SPECT™ technology, enabled high-efficiency, targeted cardiac myocardial perfusion imaging in under 5 minutes.4 Specifically, IQ•SPECT technology combines the organ-focused collimation from SMARTZOOM with an organ-centric orbit and acquisition protocol, and an iterative image reconstruction algorithm with resolution recovery that corrects for hardware-specific gantry, collimator, and system point response physics, to provide high-quality images for the targeted clinical application.4 The cardio-focused detector orbit enables the user to specify where the heart is located; and system then automatically creates a cardio-centric orbit for the acquisition. The SMARTZOOM collimator with IQ•SPECT was shown in a myriad of studies to improve imaging sensitivity by at least four times, thus allowing for cardiac perfusion imaging with half the radiotracer dose and in half the scan time.6-14 White paper · SMARTZOOM HRX 3 Expanding the SMARTZOOM product line, the SMARTZOOM HRX collimator leverages a higher-order design equation and state-of-the-art manufacturing quality control automation methods to provide a larger magnification volume with high magnification homogeneity and resolution compared to the original SMARTZOOM collimator, at a cost of diminished sensitivity.15 This larger “sweet spot,” or the region where magnifi- cation is greater than 1, means SMARTZOOM HRX is well-suited to accommodate neurological applications, as well as small FOV orthopedic and oncological applications, eg, prostates and joints.16,17 Additionally, SMARTZOOM HRX benefits from Siemens Healthineers’ IQ•Neuro technology, which combines the SMARTZOOM HRX collimator with a neuro-centric detector orbit, a scan protocol, and an advanced iterative image reconstruction algo- rithm that models the detector, system, and collimator physics, including accounting for SMARTZOOM HRX collimator’s precise hole sizes, shapes, and angular directions and all. The neuro-centric detector orbit is achieved with a 180-degree detector config- uration and circular orbit, where the user needs to ensure the organ of interest is in the center of the FOV. Furthermore, SMARTZOOM HRX was designed for use with both 99mTc and 123I radio- tracers, with the septal penetration for 123I comparable to that of medium-energy low-penetration (MELP) collimators.17 As such, this white paper reviews key pertinent information related to the entire SMARTZOOM HRX offering, including the motivation behind its creation and its intended clinical applications, and discusses its innovative engineering design as well as its performance and validation data. Clinical applications As the SPECT paradigm is expanding away from high-volume and -throughput myocardial perfusion imaging towards higher-resolution and quantitative imaging across multiple clinical indications, it became apparent a novel SMARTZOOM collimator was necessary to meet the demands of the changing field. Neurological applications The conventional protocol for neurological scans with general-purpose SPECT and SPECT/CT scanners involves positioning the detectors above the shoulders and main- taining a close-up, non-circular orbit (NCO) around the head with parallel-hole collima- tors to produce the highest resolution scans of the brain. This results in the brain being positioned in the lower portion of the detector FOV versus the center. However, the long scan duration and nearby moving detectors can contribute to patient discomfort and claustrophobia. Furthermore, pediatric neurological scans require higher-resolution images to reliably resolve their smaller brain structures and inform clinical decisions. For example, identification of small focal hyperperfused lesions during seizures with 99mTc perfusion agents can help to inform surgical planning in pediatric epilepsy.18,19 Additionally, scans with 123I-labelled tracers, such as 123I-Iofluplane (DaTscan), have higher emission energies than 99mTc scans, which cause greater septal penetration with lower-energy collimators like the low-energy high-resolution (LEHR) and degrades image quality. Therefore, the rationale for designing SMARTZOOM HRX was to develop a collimator that produces a high-resolution and high-sensitivity image of the brain from a wider and more comfortable scan orbit for both 99mTc and 123I energies. The expectation is that this collimator will improve neurological protocols from both image quality and workflow perspectives and help advance the field towards a future of fully quantitative imaging. 4 White paper · SMARTZOOM HRX The specific technical details of SMARTZOOM HRX for neurological applications are detailed in Section 4. Briefly, SMARTZOOM produces similar resolution for 99mTc imaging from a comfortable 26-cm circular acquisition orbit with xSPECT™ extra-modal (EM) reconstruction, as the LEHR, collimator does in a close-up 16.5-cm NCO with Flash 3D™ iterative reconstruction and resolution recovery.15 Additionally, SMARTZOOM HRX has a high sensitivity of 371 cpm/µCi at 28 cm compared to LEHR, which has 202 cpm/ µCi at 10 cm (In SI units: 167 cps/MBq for SMARTZOOM HRX at 28 cm vs. 91 cps/MBq for LEHR at 10 cm).20 Furthermore, for 123I, the septal penetration is similar to that of the MELP collimator.17 Example clinical neurological images can be seen in Figure 2.1.21 Figure 2.1: Example 99mTc brain perfusion scans compared between a 26-cm neuro-centric SMARTZOOM HRX acquisition and a typical NCO LEHR acquisi- tion. Both images were acquired RFP with a continuous orbit consist- ing of 60 frames, a 3-degree SZHRX LEHR angular separation, and 4. 3 20 seconds per frame. Then were subsequently recon- LEHR HAR structed with the xSPECT-EM AFR algorithm with 24 iterations and 4 subsets and a 10-mm gaussian post-smoothing filter. SZHRX The SMARTZOOM HRX scan demonstrate higher cortical LEHR HAR and basal ganglia contrast com- RFP pared to the background while operating in a wider and more comfortable orbit. SZHRX Data courtesy on file. Other clinical applications There are many diverse SPECT applications across the orthopedic and oncological fields, with the goal of acquiring images of a specific organ, lesion, tumor, or joint that comprises a small fraction of the detector FOV. While the exact acquisition details are heavily dependent upon each clinical task in question, many of these tasks could benefit from a collimator that produces a focused image of higher-resolution and higher-sensitivity from a wider orbit acquisition. As noted, neurological acquisitions, the small, targeted region must be positioned in the center of the FOV and scan orbit to benefit from SMARTZOOM HRX’s improved imaging sensitivity. White paper · SMARTZOOM HRX 5 Brain scan positioning Although SMARTZOOM HRX neurological scans begin with the detectors oriented vertically in the 0- and 180-degree positions, SMARTZOOM HRX best practices involve first performing a manual localization step to position the brain at isocenter and in the center of the detector FOV. This ensures that the brain is located in the center of rotation and within the magnification volume to benefit from the SMARTZOOM HRX’s imaging sensitivity improvements. This protocol is different from traditional neuro- logical scans performed in close-up NCO, where the detectors are placed above the shoulders, which results in the brain sitting at the bottom of the FOV. Instead, SMARTZOOM HRX brain acquisitions should position the brain in the center of the detectors as detailed below. To perform the localization, it is recommended to first position both detectors horizon- Figure 3.1: Illustration of the tally at 90- and 270- degrees (3 and 9 o’clock positions, see Figure 3.1 A,B) and at the SMARTZOOM HRX brain posi- desired 26-cm distance. Without moving the detectors, the technologist will maneuver tioning step from an example the patient and bed to place the brain in the center of the FOV based on the yellow Hoffman brain phantom scan. (A) Panel A depicts the gantry boxes in the gantry display. Figure 3.1 A,C demonstrates the ideal pre-scan localization display during the brain posi- for a Hoffman brain phantom scan performed at Siemens Healthineers’ factory. For this tioning step. The yellow arrows experiment, a table height of around 148 mm and table depth of -114 mm resulted in highlight the correct position- the proper central positioning. The values will inevitably vary across different patients, ing of the brain in the center but these numbers could serve as a rough starting point from which finer adjustments of the yellow detector boxes. The green boxes highlight the can be made by the technologist. Additionally, depending on the scanner software detectors, which are positioned version, a smaller yellow box and central crosshair will appear on the gantry to further horizontally in the 90- and assist with localization (Figure 3.1C). Once the positioning step is complete, the detec- 270-degree configuration and tors can be reset to the vertical orientation (0- and 180-degree orientation) and at the desired scan distance of commence the acquisition. 26 cm. The red box highlights the resulting table position with a height of 148 mm and depth of -114 mm. These num- Patient registered A B bers will vary across different b.8 11. patients and are only meant Brain SZ 180 SPECT-CT [factory] 1476 as approximate starting points Tomo for the technicians to perform manual localization. (B) Panel B is a picture of the scanner room during the positioning step Table.SPECT Gantry -113.5 Detector 1 -1 260 O 90.00 Detector 2: Not Tilted with the detectors in the 90- -I 260 90º -90.0º and 270-degree configuration. 147.6 = 0.0 (C) The yellow arrows in Panel C Gantry Rotation $ 0.0º O highlight the brain localization SZHRX 260 260 90 boxes and crosshairs which Patient registered o may be present depending on C 0.4 10.5 D the software version to assist Brain Sz 180 SPECT-CT [factory] with centering the acquisition. Tomo (D) Panel D displays example reconstructed images with the brain in the center of the detector FOV. Head First - Supine (HFS) 6 White paper · SMARTZOOM HRX Performance and validation SMARTZOOM HRX was created for high-resolution and high-sensitivity organ-specific imaging of 99mTc and 123I radiotracers. Resolution benchmarking SMARTZOOM HRX was designed to operate in an organ-centric circular orbit at a distance of 26–28 cm with a similar effective resolution as the LEHR collimator in a close-up 16.5-cm NCO with Flash 3D iterative reconstruction with resolution recovery.15 The resolution performance of SMARTZOOM HRX was benchmarked compared to LEHR in an experiment with the Data Spectrum cylinder phantom with hot-cold spheres (HCS) (see Figure 4.1).15 The HCS experiment allows for quantitative assessment of resolution and noise performance in tasks mimicking neurological clinical applications of 99mTc brain perfusion imaging (hot- and cold-lesion detection tasks) and 123I Ioflupane (DaTscan) (cold-lesion detection task). In this experiment, the HCS phantom was scanned with SMARTZOOM HRX in an organ-centric 28-cm orbit, as well as LEHR in a close-up 16.5-cm NCO and LEHR in a far 28-cm orbit. All images were reconstructed with xSPECT Quant™ reconstruction methodology based on a conjugate gradient optimizer and a system calibration with a 3%-traceable National Institute of Standards and Technology (NIST)-calibrated sensitivity source.22 A “reconstructed resolution” metric was used to compare the effective resolution between the three scans. This “reconstructed resolution” metric is based on the edge response of the SPECT image and a matched filter applied to the ground truth higher resolution CT image, which results in a criterion that is robust against variations in image contrast, acquisition, or reconstruction method.22-24 Graphs illustrating computed reconstructed resolution versus reconstruction update number and noise are shown in Figure 4.1. For the task of cold-sphere detection, at 48 iterations, SMARTZOOM HRX results in a resolution of 9.7 ± 0.2 mm (mean ± standard deviation) compared to LEHR at NCO resolution of 10.8 ± 0.2 mm. Similarly, for the task of hot-sphere detection, at 48 iterations SMARTZOOM HRX results in a resolution of 10.7 ± 0.1 mm compared to LEHR at NCO resolution of 10.3 ± 0.1 mm. Given the similar reconstructed resolutions between SMARTZOOM HRX at 28 cm and LEHR at NCO, we can conclude that SMARTZOOM HRX achieved its designated purpose of producing similar resolution scans to that of LEHR in close-up NCO while maintaining a wider, organ-centric orbit. While this experiment used a 28-cm orbit and xSPECT quantitative methodology, the product recommends a default 26-cm orbit and may not support quantitative imaging. However, these experimental conditions are more stringent than product defaults and would not affect the conclusions drawn from this experiment. White paper · SMARTZOOM HRX 7 Figure 4.1: Images and data 15 15 XQ SZHRX at 28 cm XQ SZHRX at 28 cm ++ XQ SZHRX at 28 cm XQ SZHRX at 28 cm from the resolution benchmark XQ LEHR at NCO XQ LEHR at NCO XQ LEHR at NCO XQ LEHR at NCO 14 XQ LEHR at 28 cm 14 XQ LEHR at 28 cm 14 XQ LEHR at 28 cm 14 XQ LEHR at 28 cm -- I experiment in the HCS phan- tom.15 SMARTZOOM HRX acqui- 13 13 13 13 sition at 28 cm was compared +-+ -- t --- to LEHR acquisitions in close-up 12 12 12 12 HERE ITHAY NCO and far circular orbit of 28 -> ---- 11 11 11 11 cm. All images were recon- Hot spheres resolution (mm) Hot spheres resolution (mm) Cold spheres resolution (mm) Cold spheres resolution (mm) structed with the conjugate 10 10 10 10 gradient method with scatter 9 and CT attenuation correction. 20 40 60 80 Reconstruction Updates 100 20 40 60 80 0.10 Reconstruction Updates 100 0.05 0.15 0.20 0.25 0.05 0.10 0.15 0.20 0.25 Noise olu Noise olu Curves displaying the computed reconstruction resolution22,23 versus reconstruction update CT SZHRX LEHR LEHR (top left) and versus noise (top 28 cm NCO 28 cm 40 right) are displayed for both hot- and cold-sphere detection tasks. The error bars show one Cold 30 spheres standard deviation in calculated noise and resolution from 10 20 different noise realizations. On the bottom, CT images are 10 displayed along with the corre- sponding 3 SPECT acquisitions Hot for the hot- and cold-sphere spheres 0 slices. SMARTZOOM HRX recon- kBq/mL struction at 26 cm appears to resemble LEHR at NCO. Validation data for neurological applications For neurological applications, SMARTZOOM HRX was designed for high sensitivity and wider acquisitions of 99mTc and 123I radiotracers, with a larger magnification volume to better encompass the whole brain.17 The larger 26–28-cm orbit is intended to be large enough to clear most adult shoulders while improving patient comfort and reducing anxiety.17 The performance of the SMARTZOOM HRX collimator for neurological perfusion applications was evaluated in two studies of the Hoffman brain phantom.17,25 The first study investigated the efficacy of SMARTZOOM HRX in producing clinical contrasts for 99mTc brain perfusion cold detection tasks.25 SMARTZOOM HRX scans acquired at 26 and 28 cm and reconstructed with the ordered subset conjugate gradient minimizer (OSCGM) reconstruction algorithm were compared to LEHR acquisitions at 15-, 26-, and 28-cm orbits reconstructed with Flash 3D (see Figure 4.2). Qualitatively, the SMARTZOOM HRX images at distant orbits of 26-28 cm produced high-quality images, which were superior to LEHR images at the same distances and more closely resembled the LEHR at 15 cm. Quantitative image comparisons were drawn by evaluating the percent contrast (% contrast) between the grey matter (GM) and white matter (WM) regions of the phantom, the % contrast between the WM and the thalamus, and by computing the coefficient of variation in the clinically important regions of the thalamus and the cerebellum. SMARTZOOM HRX had higher GM-WM % contrast at distances of 26 and 28 cm than the LEHR at distances of 15, 26, and 28 cm (41.4 and 43.1% vs. 33.1, 33.7, and 31.5%, respectively). 8 White paper · SMARTZOOM HRX SMARTZOOM HRX also exhibited greater homogeneity within the right thalamus at 26 and 28 cm compared to the LEHR at 15, 26, and 28 cm (SMARTZOOM HRX: 8.4 and 8.6% vs. LEHR: 13.3, 10.5, and 10.0%), as well as in the left thalamus (SMARTZOOM HRX: 7.5 and 8.1% vs. LEHR: 13.8, 12.6, and 11.1%), and in the cerebellum (SMARTZOOM HRX: 8.3 and 7.3% vs. LEHR: 9.2, 13.0, and 11.6%). However, while SMARTZOOM HRX acquisitions at 26 and 28 cm had higher % contrast between the WM and thalamus compared to LEHR at the same distances, the LEHR at 15 cm produced the highest % contrast for the WM and thalamus (SMARTZOOM HRX: 8.3 and 7.7% vs. LEHR: 24.4, 0.7 and 1.7%). Overall, the conclusion of this study was SMARTZOOM HRX produces similar image quality to LEHR while operating in more comfortable distant orbits.25 The second study investigated the feasibility of quantitative 99mTc brain perfusion imaging in the Hoffman brain phantom.17 Absolute quantitative activity concentrations in kBq/mL were obtained with the xSPECT-EM method22 for SMARTZOOM HRX in 28-cm orbit as well as for the LEHR in NCO and at 28 cm. Images were assessed qualitatively in addition to quantitative analyses comparing the imaging sensitivity and the measured activity concentrations in the whole WM region, the whole GM region, as well as ran- domly sampled small spheres within each tissue region. Qualitative image analysis suggested SMARTZOOM HRX at 28 cm produced better imaging resolution than LEHR at 28 cm and that the resolution was closer to LEHR at NCO. Quantitative image analy- ses showed no significant difference in measured quantitative activity concentrations between SMARTZOOM HRX and LEHR in the GM or WM regions. Furthermore, the SMARTZOOM HRX at 28 cm was shown to have high imaging sensitivity compared to LEHR at the same distance with an improvement of 1.38x as many counts.17 While this experiment used a 28-cm orbit and xSPECT quantitative methodology, the product recommends a default 26-cm orbit and may not support quantitative imaging. However, these experimental conditions are more stringent than product defaults and would not affect the conclusions drawn from this experiment. Figure 4.2: Brain perfusion 99mTc scans of the Hoffman brain phantom compared LEHR SMARTZOOM HRX between the SMARTZOOM HRX and LEHR collimators at differ- 15 cm 26 cm 28 cm 26 cm 28 cm ent detector orbit distances.25 SMARTZOOM at far scan orbits of 26/28 cm can be seen to have better resolution than LEHR at the same distances and more closely resembles that of the LEHR at close-up orbit of 15 cm. Data courtesy of the University of Kanazawa, Kanazawa, Japan.25 White paper · SMARTZOOM HRX 9 Collimator design and image reconstruction Similar to IQ•SPECT technology, IQ•Neuro is the combination of the SMARTZOOM HRX multifocal collimator and a neuro-centric acquisition mode and reconstruction engine for faster neurological scanning (Figure 5.1). Unique magnifying collimators Neuro-centric acquisition Advanced reconstruction SMARTZOOM HRX Healthingers Figure 5.1: Overview of IQ•Neuro technology. Design SMARTZOOM HRX differs from the cardiac SMARTZOOM collimator in that it leverages smaller 1.9-mm hexagonal tapered pins, a higher order design equation, and adapted focal lengths to accommodate a larger and more homogenous “sweet spot” or magnification area (see Figures 5.2, 5.3).15 Figure 5.2 demonstrates SMARTZOOM HRX’s larger magnification area compared to the SMARTZOOM collimator over a wide range of detector distances. Figure 5.3 illustrates the magnification area and uniformity of both SMARTZOOM HRX and SMARTZOOM multifocal collimators at the standard clinical operating distances of 26–28 cm. At a 26-cm detector distance, SMARTZOOM HRX has a planar magnification area of 52,627 mm2 with a magnification standard deviation of 0.74, whereas SMARTZOOM has a smaller magnification area of 36,139 mm2 with a magnification standard deviation of 1.56. The increase in magnification area and improvement in magnification homogeneity was designed to allow SMARTZOOM HRX to fit around the adult brain and maintain uniform clinical contrast across the focusing region.17 However, it is important to note that the average and peak magnification in SMARTZOOM HRX is lower than in the cardiac-only SMARTZOOM collimator, and therefore the realized improvement in sensitivity or scan time will be lower than that reported for the SMARTZOOM collimator. For example, at the 26-cm distance shown in Figure 5.3, SMARTZOOM HRX has an average magnification within the sweet spot of 3.78, compared to the SMARTZOOM collimator, which has a greater average magnification of 4.23. Per NEMA NU-1 2018 standards, SMARTZOOM HRX has a lower sensitivity of 371 cpm/µCi (167 cps/MBq) at 28 cm compared to 810 cpm/µCi (364 cps/MBq) from the SMARTZOOM collimator at the same distance.20 10 White paper · SMARTZOOM HRX Automated quality control for all manufactured collimators Like SMARTZOOM collimators, all SMARTZOOM HRX collimators are fully characterized in the factory prior to shipping. Since there are always inevitable minor variations in col- limator hole sizes and angular directions that arise during the casting of the collimators, Siemens Healthineers uses a proprietary method to measure and save each collimator’s physical properties prior to shipping.4 Then, this high-resolution collimator optical vector map is used in each iterative image reconstruction algorithm to account for and correct for the slight deviations and deliver high-quality data to the user. Organ-centric acquisition The SMARTZOOM HRX collimator is designed to operate in a wide, circular orbit at a 26–28-cm radius (Figure 5.1).15 For optimal sensitivity gain, the orbit is carefully chosen such that the organ of interest remains within the magnification volume throughout the entirety of the scan (Figure 3.1).4 For neurological applications, the 26–28-cm radius is large enough to clear most shoulders and can therefore improve patient comfort and can reduce claustrophobia and anxiety.17 This also means the brain is situated more cen- trally in the detector FOV than in typical neurological SPECT scans, where the detectors must be positioned more cranially to avoid contact with the shoulders. Furthermore, the system-specific 3D gantry deflection correction obviates the need for previously used head-alignment calibrations or verification for SMARTZOOM HRX. Image reconstruction SMARTZOOM HRX images are generated with an iterative image reconstruction approach based on the expectation maximization or conjugate gradient optimizers.22 As stated above, the reconstruction method strives to model all important physical phenomena that would impact the acquired data and diminish image resolution. This includes measuring all gravitational deflections and deviations in the gantry’s position and orientation that result from every conceivable detector configuration, orbit orienta- tion, radial distance, and SMARTZOOM HRX collimator weights, and then applying the corresponding affine transformations to correct the data in image space. Additionally, both reconstruction algorithms use each SMARTZOOM HRX collimator’s specific hole sizes and angular directions that were measured after fabrication to correct the projec- tion operator and improve the resultant image quality. Furthermore, an accurate system point response function (PRF) is produced by convolving the detector’s native Gaussian PSF with SMARTZOOM HRX’s hexagonal aperture geometry to generate a composite system “hex-cone” PSF. Moreover, the iterative reconstruction uses Mighell’s modified Chi-squared-gamma merit function to account for Poisson statistics. Finally, attenuation and scatter correction are enabled in the system matrix and the forward projection oper- ator, respectively. This entire approach adopts the “data is sacred” philosophy, wherein the data itself is not modified, but instead all the system physics corrections are applied in image space.4 White paper · SMARTZOOM HRX 11 Figure 5.2: Planar magnification compared between SMARTZOOM and SMARTZOOM HRX multifocal SMARTZOOM SMARTZOOM HRX collimators as a function of distance. Magnification areas 5 cm 10 cn 15 cn 5 cm 10 cn 15 cm are computed in planes parallel to the collimator surface at orthogonal distances of 5 cm to 45 cm. SMARTZOOM HRX has a larger magnification area at each distance, which allows it 20 cm 25 cm 30 cm 20 cm 25 cm 30 cm to target larger organs such as the brain. The data presented here are based on high-resolu- tion optical measurements of each individual collimator hole 35 cm 40 cm 45 cm 35 cm 40 cm 45 cm size and orientation. Addition- ally, the display windowing is different for each planar dis- tance to enable comparisons of the magnification areas between the two collimators. SMARTZOOM SMARTZOOM HRX 26 cm Figure 5.3: Planar magnification compared Amag = 36139 mm2 Amag = 52627 mm2 between SMARTZOOM and H, O mag = 4.23 + 1.56 H, O mag = 3.78 + 0.74 SMARTZOOM HRX multifocal collimators. Magnification areas Amag, average values μmag, and standard deviations σmag are computed in planes parallel to the collimator surface at oper- ating distances of 26–28 cm. SMARTZOOM HRX has a larger 28 cm focusing area with greater magnification homogeneity but lower average magnification compared to SMARTZOOM Amag = 34287 mm2 Amag = 50085 mm2 collimator. The larger area and greater homogeneity allow H, O mag = 5.07 + 1.96 H, O mag = 4.38 + 0.85 SMARTZOOM HRX to maintain uniform image contrast while targeting larger organs such as Magnification the brain. The data presented here are based on high-resolu- tion optical measurements of 0 1 2 3 4 5 6 7 8 each individual collimator hole size and orientation. 12 White paper · SMARTZOOM HRX SMARTZOOM HRX considerations Semi-quantitative defect scores are known to aide clinicians in reading and interpreting SPECT images, and this is especially true for acquisitions with multi-focal collimators.5,7,26 Defect scores are obtained by referencing the patient scan to a “normal database,” or a database of healthy acquisitions that represent the normal physiologic variations in radiotracer uptake and match the image conditions and agents used. Normal data- bases for SMARTZOOM HRX applications are under development for each radiotracer and application and not available currently. Users are encouraged to create their own normal databases. About the lead author Brandon Clinton Jones, PhD, is a research scientist focused on improving medical imaging technologies through the use of quantitative imaging, artificial intelligence, and computational physics. His expertise includes designing novel quantitative imaging protocols, developing machine learning approaches to improve image quality and mitigate artifacts, optimizing image acquisition/reconstruction algorithms, and running clinical studies in human participants. Brandon earned a PhD in bioengineering and a master’s degree in scientific computing from the University of Pennsylvania, Philadelphia, Pennsylvania. White paper · SMARTZOOM HRX 13 References 1 Rajaram R, et al. Tomographic performance characteristics of the IQ•SPECT system. IEEE Medical Imaging Conference Record. 2011. 2 Zeintl J, et al. Performance characteristics of the SMARTZOOM collimator. IEEE Medical Imaging Conference Record. 2011. 3 Zeintl J, et al. First experience with SMARTZOOM collimation in clinical cardiac SPECT. Annual Congress of the European Association of Nuclear Medicine. 2009. Barcelona, Spain. 4 Vija AH, et al. A method for improving the efficiency of myocardial perfusion imaging using conventional SPECT and SPECT/CT imaging systems. Nuclear Science Symposium Conference Record (NSS/MIC). 2010 IEEE. 2010. 5 Nakajima K, et al. IQ•SPECT technology and its clinical applications using multicenter normal databases. Ann Nucl Med, 2017. 31(9): 649-659. 6 Hyafil F, et al. EANM procedural guidelines for myocardial perfusion scintigraphy using cardiac-centered gamma cameras. European Journal of Hybrid Imaging. 2019. 3(1): 11. 7 Konishi T, et al. IQ•SPECT for thallium-201 myocardial perfusion imaging: effect of normal databases on quantification. Annals of Nuclear Medicine. 2017. 31(6): 454-461. 8 Lyon MC, et al. Dose reduction in half-time myocardial perfusion SPECT-CT with multifocal collimation. Journal of Nuclear Cardiology. 2016. 9 Gremillet E, Agostini D. 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. 2015. 43(4): 707-710. 10 Caobelli F, et al. Feasibility of one-eighth time gated myocardial perfusion SPECT functional imaging using IQ•SPECT. Eur J Nucl Med Mol Imaging. 2015. 42(12): 1920-1928. 11 Caobelli F, et al. IQ•SPECT allows a significant reduction in administered dose and acquisition time for myocardial perfusion imaging: evidence from a phantom study. Journal of Nuclear Medicine. 2014. 55: 2064-2070. 12 Horiguchi Y, et al. Validation of a short-scan-time imaging protocol for thallium-201 myocardial SPECT with a multifocal collimator. Annals of Nuclear Medicine. 2014: 1-9. 13 Caobelli F, et al. Evaluation of patients with coronary artery disease. Nuklearmedizin. 2013. 52. 14 Imbert L, et al. Compared performance of high-sensitivity cameras dedicated to myocardial perfusion spect: a comprehensive analysis of phantom and human images. Journal of Nuclear Medicine. 2012. 53: 1897-1903. 14 White paper · SMARTZOOM HRX 15 Massanés F, Vija AH. Characterization of a high-resolution multi-focal SPECT collimator. 2021 IEEE Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC). 2021. 16 Massanes F, et al. New multi-focal collimator designed for quantitative SPECT imaging in nuclear cardiology applications. SNMMI 2021. 2021. 17 Massanes-Basi F, et al. New multi-focal collimator designed for quantitative SPECT imaging in nuclear neurology applications. European Association of Nuclear Medicine 21st Annual Congress. 2021: Virtual. 18 Kuruva M, Moncayo VM, Peterson RB. PET and SPECT imaging of epilepsy: technical considerations, pathologies, and pitfalls. Seminars in Ultrasound, CT and MRI. 2020. 41(6): 551-561. 19 Ergün EL, et al. SPECT-PET in epilepsy and clinical approach in evaluation. Semin Nucl Med. 2016. 46(4): 294-307. 20 Siemens Healthineers, System specifications for Symbia Pro.specta SPECT/CT in Molecular Imaging SharePoint. 2024. 21 Kamada S, et al. Operation and clinical benefits of the SPECT/CT system Symbia pro.specta. Eizo Joho Medical (Japan). 2023. 55(11): 22-30. 22 Vija AH. Characteristics of the xSPECT reconstruction method. 2017. Siemens Medical Solutions USA, Inc., Molecular Imaging. 23 Ma J, Vija AH. Comparison of multi modal SPECT reconstruction methods using a clinically relevant assessment of the image resolution at various noise levels. IEEE Medical Imaging Conference Record. 2014. 24 Fessler JA, Rogers WL. Spatial resolution properties of penalized-likelihood image reconstruction: space-invariant tomographs. IEEE Transactions on Image Processing. 1996. 5(9): 1346-1358. 25 Shibutani T, et al. Image characteristics of brain perfusion SPECT/CT using a new multi-focal collimator: Comparison with conventional SPECT with LEHR collimator. Journal of Nuclear Medicine. 2022. 63(supplement 2): 4116. 26 Okuda K, et al. Creation and characterization of normal myocardial perfusion imaging databases using the IQ•SPECT system. Journal of Nuclear Cardiology. 2018. 25(4): 1328-1337. White paper · SMARTZOOM HRX 15 Trademarks and service marks used in this material SMARTZOOM HRX is exclusively offered on are property of Siemens Medical Solutions USA or Symbia Pro.specta™ SPECT/CT systems. Siemens Healthineers AG. All other company, brand, product, and service names may be trademarks or Symbia Pro.specta and its features are not commercially registered trademarks of their respective holders. available in all countries. Future availability cannot be Please contact your local Siemens Healthineers sales guaranteed. representative for the most current information or contact one of the addresses listed below. All comparative claims derived from competitive data at the time of printing. Data on file. Siemens Healthineers reserves the right to modify the design and specifications contained herein without prior notice. As is generally true for technical specifications, the data contained herein varies within defined tolerances. Some configurations are optional. Product performance depends on the choice of system configuration. Note: Original images always lose a certain amount of detail when reproduced. “Siemens Healthineers” is considered a brand name. Its use is not intended to represent the legal entity to which this product is registered. All photographs © 2024 Siemens Healthineers AG. All rights reserved. Siemens Healthineers Headquarters Published by Siemens Healthineers AG Siemens Medical Solutions USA, Inc. Siemensstr. 3 Molecular Imaging 91301 Forchheim 2501 N. Barrington Road Germany Hoffman Estates, IL 60192 Phone: +49 9191 18-0 USA siemens-healthineers.com Phone: +1 847 304-7700 siemens-healthineers.com/mi MI-6601 TA.JV ∙ PDF ONLY ∙ © Siemens Healthineers AG, 11.2024