Correction of partial volume effects in myocardial SPECT

A Deep-Learning-Based Partial-Volume Correction Method for Quantitative 177Lu SPECT/CT Imaging

With the development of new radiopharmaceutical therapies, quantitative SPECT/CT has progressively emerged as a crucial tool for dosimetry. One major obstacle of SPECT is its poor resolution, which results in blurring of the activity distribution. Especially for small objects, this so-called partial-volume effect limits the accuracy of activity quantification. Numerous methods for partial-volume correction (PVC) have been proposed, but most methods have the disadvantage of assuming a spatially invariant resolution of the imaging system, which does not hold for SPECT. Furthermore, most methods require a segmentation based on anatomic information. Methods: We introduce DL-PVC, a methodology for PVC of 177Lu SPECT/CT imaging using deep learning (DL). Training was based on a dataset of 10,000 random activity distributions placed in extended cardiac-torso body phantoms. Realistic SPECT acquisitions were created using the SIMIND Monte Carlo simulation program. SPECT reconstructions without and with resolution modeling were performed using the CASToR and STIR reconstruction software, respectively. The pairs of ground-truth activity distributions and simulated SPECT images were used for training various U-Nets. Quantitative analysis of the performance of these U-Nets was based on metrics such as the structural similarity index measure or normalized root-mean-square error, but also on volume activity accuracy, a new metric that describes the fraction of voxels in which the determined activity concentration deviates from the true activity concentration by less than a certain margin. On the basis of this analysis, the optimal parameters for normalization, input size, and network architecture were identified. Results: Our simulation-based analysis revealed that DL-PVC (0.95/7.8%/35.8% for structural similarity index measure/normalized root-mean-square error/volume activity accuracy) outperforms SPECT without PVC (0.89/10.4%/12.1%) and after iterative Yang PVC (0.94/8.6%/15.1%). Additionally, we validated DL-PVC on 177Lu SPECT/CT measurements of 3-dimensionally printed phantoms of different geometries. Although DL-PVC showed activity recovery similar to that of the iterative Yang method, no segmentation was required. In addition, DL-PVC was able to correct other image artifacts such as Gibbs ringing, making it clearly superior at the voxel level. Conclusion: In this work, we demonstrate the added value of DL-PVC for quantitative 177Lu SPECT/CT. Our analysis validates the functionality of DL-PVC and paves the way for future deployment on clinical image data.

A new cardiac phantom for dynamic SPECT

BACKGROUND

In recent years, with the advance of myocardial blood flow (MBF) measurement capability in dynamic single photon emission computerized tomography (SPECT) systems, significant effort has been devoted to validation of the new capability. Unfortunately, the mechanical phantoms available for the validation process lack essential features-they either have a constant radiotracer concentration or they have rigid (static) walls unable to simulate cardiac beating.

METHODS AND RESULTS

We have developed a mechanical cardiac phantom that is able to mimic physiological radiotracer variation in the left ventricle (LV) cavity and in the myocardium (M), while performing beating-like motion. We have also developed a mathematical model of the phantom, allowing a description of the radiotracer concentrations in both regions (LV, M) as a function of time, which served as a tool for experiment planning and to accurately mimic physiological-like time-activity curves (TACs). A net retention model for the phantom was also developed, which served to compute the theoretical (i.e., expected) MBF of the phantom from measured quantities only, and thus validate the MBF reported by the SPECT system. In this paper, phantom experiments were performed on a GE Discovery NM 530c SPECT system.

CONCLUSIONS

A novel dynamic cardiac phantom for emission tomography has been developed. The new phantom is capable of producing a wide range of TACs that can mimic physiological (and potentially in the future, pathological) curves, similar to those observed in dynamic SPECT systems. SPECT-reported MBF values were validated against known (measured) activity of the injected radiotracer from phantom experiments, which allowed to determine the accuracy of the GE Discovery 530c SPECT system.

Assessment of cardiac amyloidosis with 99mTc-pyrophosphate (PYP) quantitative SPECT

Background

^99mTc-PYP scintigraphy provides differential diagnosis of ATTR cardiomyopathy (ATTR-CM) from light chain cardiac amyloidosis and other myocardial disorders without biopsy. This study was aimed to assess the diagnostic feasibility and the operator reproducibility of ^99mTc-PYP quantitative SPECT.

Method

Thirty-seven consecutive patients who underwent a ^99mTc-PYP thorax planar scan followed by SPECT and CT scans to diagnose suspected ATTR-CM were enrolled. For the quantitative SPECT, phantom studies were initially performed to determine the image conversion factor (ICF) and partial volume correction (PVC) factor to recover ^99mTc-PYP activity concentration in the myocardium for calculating the standardized uptake value (SUV) (unit: g/ml). SUV_max was compared among groups of ATTR-CM, AL cardiac amyloidosis, and other pathogens (others) and among categories of Perugini visual scores (grades 0–3). The intra- and inter-operator reproducibility of quantitative SPECT was verified, and the corresponded repeatability coefficient (RPC) was calculated.

Results

The ICF was 79,327 Bq/ml to convert count rate in pixel to ^99mTc activity concentration. PVC factor as a function of the measured activity concentration ratio in the myocardium and blood-pool was [y = 1.424 × (1 − exp(− 0.759 × x)) + 0.104]. SUV_max of ATTR-CM (7.50 ± 2.68) was significantly higher than those of AL (1.96 ± 0.35) and others (2.00 ± 0.74) (all p < 0.05). SUV_max of grade 3 (8.95 ± 1.89) and grade 2 (4.71 ± 0.23) were also significantly higher than those of grade 1 (1.92 ± 0.31) and grade 0 (1.59 ± 0.39) (all p < 0.05). Correlation coefficient (R²) of SUV_max reached 0.966 to 0.978 with only small systematic difference (intra = − 0.14; inter = − 0.23) between two repeated measurements. Intra- and inter-operator RPCs were 0.688 and 0.877.

Conclusions

^99mTc-PYP quantitative SPECT integrated with adjustable PVC factors is feasible to quantitatively and objectively assess the burden of cardiac amyloidosis for diagnosis of ATTR-CM.

Technical Note: Development of an ischemic defect model insert attachable to a commercially available myocardial phantom

OBJECTIVE

The purpose of this study was to develop a novel myocardial phantom insert model that attaches to commercially available myocardial phantoms and simulates an ischemic area, using three-dimensional printing technology.

METHODS

Ischemic inserts were designed to give four levels of absolute percent contrast (Low; 10%, Medium; 20%, High; 35%, and Defect; 100%) using CT images and computer-aided design software. The ischemic insert was composed of multiple slit structures to replicate myocardial ischemia. Myocardial phantom images with developed ischemic inserts were acquired using a SPECT/CT system and were then reconstructed using filtered back projection (FBP) and iterative reconstruction (IR) with various cutoff frequencies of a Butterworth filter. The performance and utility of ischemic inserts were evaluated according to percent contrast and 5-point scoring.

RESULTS

The percent contrast and scoring results changed according to the ischemic insert type, cutoff frequency, and reconstruction method. The percent contrast of each insert obtained by FBP with 0.4 cycles/cm was 4.1% (Low), 15.7% (Medium), 17.4% (High), and 36.1% (Defect). Similarly, the percent contrast of each insert obtained by IR with 0.4 cycles/cm was 5.0% (Low), 17.0% (Medium), 21.9% (High), and 47.7% (Defect).

CONCLUSIONS

We successfully developed an ischemic insert that attaches to a commercially available myocardial phantom by using CT imaging and 3D printing technology. Our proposed ischemic insert provided several abnormal perfusion patterns on myocardial SPECT images and may be useful for evaluating SPECT image quality.

Enhancement of Partial Volume Correction in MR-Guided PET Image Reconstruction by Using MRI Voxel Sizes

Positron emission tomography (PET) suffers from poor spatial resolution which results in quantitative bias when evaluating the radiotracer uptake in small anatomical regions, such as the striatum in the brain which is of importance in this paper of neurodegenerative diseases. These partial volume effects need to be compensated for by employing partial volume correction (PVC) methods in order to achieve quantitatively accurate images. Two important PVC methods applied during the reconstruction are resolution modeling, which suffers from Gibbs artifacts, and penalized likelihood using anatomical priors. The introduction of clinical simultaneous PET-MR scanners has attracted new attention for the latter methods and brought new opportunities to use MRI information to assist PET image reconstruction in order to improve image quality. In this context, MR images are usually down-sampled to the PET resolution before being used in MR-guided PET reconstruction. However, the reconstruction of PET images using the MRI voxel size could achieve a better utilization of the high resolution anatomical information and improve the PVC obtained with these methods. In this paper, we evaluate the importance of the use of MRI voxel sizes when reconstructing PET images with MR-guided maximum a posteriori (MAP) methods, specifically the modified Bowsher method. We also propose a method to avoid the artifacts that arise when PET reconstructions are performed in a higher resolution matrix than the standard for a given scanner. The MR-guided MAP reconstructions were implemented with a modified Lange prior that included anatomical information with the Bowsher method. The methods were evaluated with and without resolution modeling for simulated and real brain data. We show that the use of the MRI voxel sizes when reconstructing PET images with MR-guided MAP enhances PVC by improving the contrast and reducing the bias in six different regions of the brain using regional metrics for a single simulated data set and ensemble metrics for ten noise realizations. Similar results were obtained for real data, where a good enhancement of the contrast was achieved. The combination of MR-guided MAP reconstruction with point-spread function modeling and MRI voxel sizes proved to be an attractive method to achieve considerable enhancement of PVC, while reducing and controlling the noise level and Gibbs artifacts.

SPECT/CT: an update on technological developments and clinical applications

Functional nuclear medicine imaging with single-photon emission CT (SPECT) in combination with anatomical CT has been commercially available since the beginning of this century. The combination of the two modalities has improved both the sensitivity and specificity of many clinical applications and CT in conjunction with SPECT that allows for spatial overlay of the SPECT data on good anatomy images. Introduction of diagnostic CT units as part of the SPECT/CT system has also potentially allowed for a more cost-efficient use of the equipment. Most of the SPECT systems available are based on the well-known Anger camera principle with NaI(Tl) as a scintillation material, parallel-hole collimators and multiple photomultiplier tubes, which, from the centroid of the scintillation light, determine the position of an event. Recently, solid-state detectors using cadmium-zinc-telluride became available and clinical SPECT cameras employing multiple pinhole collimators have been developed and introduced in the market. However, even if new systems become available with better hardware, the SPECT reconstruction will still be affected by photon attenuation and scatter and collimator response. Compensation for these effects is needed even for qualitative studies to avoid artefacts leading to false positives. This review highlights the recent progress for both new SPECT cameras systems as well as for various data-processing and compensation methods.

Clinical Utility of Enhanced Relative Activity Recovery on Systolic Myocardial Perfusion SPECT: Lessons from PET

SPECT and PET myocardial perfusion images show greater myocardial intensity and homogeneity in systole than diastole because of greater systolic myocardial thickness, less partial volume loss, and enhanced activity recovery. Consequently, conventional myocardial perfusion images obtained from whole cardiac cycles have lower myocardial intensity and greater heterogeneity than systolic images. Considering relative activity distribution on SPECT systolic images may add clinical utility to whole-cycle images and wall motion.

METHODS

Patients undergoing coronary angiogram within 4 mo after SPECT myocardial perfusion imaging were reviewed. Images were interpreted by 2 masked interpreters using a 17-segment, 5-point scale to determine summed rest scores (SSS), summed stress scores, and summed difference scores on conventional and systolic images in 603 patients (55.6% no coronary artery disease [no-CAD] and 44.4% CAD). Studies were considered normal when the SSS was less than 4 and summed difference score was less than 2.

RESULTS

In the no-CAD group, systolic SSS was lower than SSS from conventional images (2 ± 2.3 vs. 3 ± 2.6, P < 0.001). In contrast, SSS derived from systolic and conventional images were not different in the obstructive CAD group (9.1 ± 7.6 vs. 9.2 ± 7.4, P = 0.559). When systolic images were considered, true-negative studies increased from 27.2% to 43.3% (P < 0.001) whereas false-positive studies decreased from 28.4% to 12.3% (P < 0.001). True-positive (38% vs. 37.2%, P = 0.505) and false-negative studies (6.5% vs. 7%, P = 0.450) were not significantly changed. Diagnostic accuracy increased from 65.2% to 80.8% (P < 0.001).

CONCLUSION

For gated SPECT myocardial perfusion imaging, when relative activity distribution on systolic images was considered, false-positive studies were reduced and diagnostic accuracy was improved.

[The Impact That SPECT Collection Angle and Collection Orbit Gives to an Image: Myocardial Digital Phantom Study]

PURPOSE

We evaluated the effect that collection angle and collection orbit condition gave to an image quantitatively by simulating the single photon emission computed tomography (SPECT) system.

METHOD

Using the Software Package of the Nuclear Medicine Data Processor for Research, we performed making of the myocardial digital phantom, three ways of different simulation of the collection angle and collection orbit, and making of reference of the uniform picture element level. We calculated NMSE for uniformity evaluation and calculated myocardial thickness full width at half maximum (FWHM) for a spatial resolution evaluation.

RESULTS

360 degrees circular orbit collection had best uniformity. 180 degrees noncircular orbit collection had best spatial resolution.

CONCLUSION

By using the digital phantom, we focused on only collection angle and collection orbit condition, and focused on two indexes of the uniformity and the spatial resolution and were able to show a quantitative index.

New approach to quantification of molecularly targeted radiotracer uptake from hybrid cardiac SPECT/CT: methodology and validation

Quantification of molecularly targeted radiotracer uptake in the myocardium from SPECT remains challenging in part due to potentially low levels of focal tracer uptake of presently available molecularly targeted agents and further degradation of cardiac SPECT by extracardiac radioactivity and partial-volume effect. The purpose of this study was to derive and validate a new SPECT quantification method for assessments of absolute radiotracer uptake in the myocardium.

METHODS

The method was integrated with a hybrid micro-SPECT/CT imaging protocol to calculate radiotracer uptake of a molecularly targeted agent in the ischemic myocardium. CT coregistered with SPECT was used to identify the position and orientation of the left ventricle. Corrections for extracardiac activity and partial-volume errors were performed via a heuristic method derived with a total count sampling scheme. Myocardial radiotracer uptake was quantified from SPECT using an external point source as a known reference. Methods were validated using an ischemic rat model injected with a (99m)Tc-labeled SPECT radiotracer targeted at αvβ3 integrin. SPECT-quantified myocardial radiotracer uptake was compared with postmortem myocardial tissue well-counted radioactivity.

RESULTS

Initial correlation between SPECT-quantified and well-counted radioactivity was fair (R(2) = 0.19, y = 0.50x + 0.05, P = 0.06) when no correction was applied to SPECT quantification. Correlation was significantly improved with tissue weight correction (R(2) = 0.84, y = 1.82x - 0.01, P < 0.001), and a trend toward the improvement of correlation was observed with extracardiac activity correction (R(2) = 0.85, y = 1.54x - 0.01, P < 0.001) and partial-volume correction (R(2) = 0.86, y = 1.68x - 0.01, P < 0.001). Reproducibility of the SPECT quantification was excellent, either with no correction (R(2) = 0.99, y = 1.00x + 0.00, P < 0.001) or with all corrections (R(2) = 1.00, y = 1.00x - 0.00, P < 0.001).

CONCLUSION

Corrections for the myocardial tissue weight, extracardiac activity, and partial-volume errors are crucial for precise assessments of myocardial radiotracer uptake using micro-SPECT/CT. The quantitative SPECT/CT approach developed provides a reasonable and reproducible in vivo estimation of absolute radiotracer uptake in a model of myocardial injury and should permit quantitative serial monitoring of subtle changes in the myocardial uptake of targeted radiotracers.

Longitudinal Evaluation of Fatty Acid Metabolism in Normal and Spontaneously Hypertensive Rat Hearts with Dynamic MicroSPECT Imaging

The goal of this project is to develop radionuclide molecular imaging technologies using a clinical pinhole SPECT/CT scanner to quantify changes in cardiac metabolism using the spontaneously hypertensive rat (SHR) as a model of hypertensive-related pathophysiology. This paper quantitatively compares fatty acid metabolism in hearts of SHR and Wistar-Kyoto normal rats as a function of age and thereby tracks physiological changes associated with the onset and progression of heart failure in the SHR model. The fatty acid analog, ¹²³I-labeled BMIPP, was used in longitudinal metabolic pinhole SPECT imaging studies performed every seven months for 21 months. The uniqueness of this project is the development of techniques for estimating the blood input function from projection data acquired by a slowly rotating camera that is imaging fast circulation and the quantification of the kinetics of ¹²³I-BMIPP by fitting compartmental models to the blood and tissue time-activity curves.

Diminishing the impact of the partial volume effect in cardiac SPECT perfusion imaging

The partial volume effect (PVE) significantly restricts the absolute quantification of regional myocardial uptake and thereby limits the accuracy of absolute measurement of blood flow and coronary flow reserve by SPECT. The template-projection-reconstruction method has been previously developed for PVE compensation. This method assumes the availability of coregistered high-spatial resolution anatomical information as is now becoming available with commercial dual-modality imaging systems such as SPECT/CTs. The objective of this investigation was to determine the extent to which the impact of the PVE on cardiac perfusion SPECT imaging can be diminished if coregistered high-spatial resolution anatomical information is available. For this investigation the authors introduced an additional parameter into the template-projection-reconstruction compensation equation called the voxel filling fraction (F). This parameter specifies the extent to which structure edge voxels in the emission reconstruction are filled by the structure in question as determined by the higher spatial-resolution imaging modality and the fractional presence of the structure at different states of physiological motion as in combining phases of cardiac motion. During correction the removal of spillover to the cardiac region from the surrounding structures is performed first by using reconstructed templates of neighboring structures (liver, blood pool, lungs) to calculate spillover fractions. This is followed by determining recovery coefficients for all voxels within the heart wall from the reconstruction of the template projections of the left and right ventricles (LV and RV). The emission data are subsequently divided by these recovery coefficients taking into account the filling fraction F. The mathematical cardiac torso phantom was used for investigation correction of PVE for a normal LV distribution, a defect in the inferior wall, and a defect in the anterior wall. PVE correction resulted in a dramatic visual reduction in the impact of extracardiac activity, improved the uniformity of the normally perfused heart wall, and enhanced defect visibility without undue noise amplification. No significant artifacts were seen with PVE correction in the presence of mild (one voxel) misregistration. A statistically significant improvement in the accuracy of the count levels within the normal heart wall was also noted. However, residual spillover of counts from within the myocardium creates a bias in regions of decreased wall counts (perfusion defects/abnormal wall motion) when the anatomical imaging modality does not allow definition of templates for defects present in the heart during emission imaging.

Left bundle-branch block artifact on single photon emission computed tomography with technetium Tc 99m (Tc-99m) agents: mechanisms and a method to decrease false-positive interpretations

Myocardial perfusion scintigraphy is a well validated noninvasive method of evaluating for significant coronary artery disease, especially in cases where electrocardiographic changes are nondiagnostic, including left bundle-branch block. However, such testing with a technetium Tc 99m agent is often confounded by left ventricular septal-based false-positive perfusion defects. These defects can be either reversible or irreversible in the septal or anteroseptal wall, problematically then, in the territory supplied by the left anterior descending coronary artery. Mechanisms explaining false-positive defects include decreased perfusion via impaired microvessel flow and normal perfusion with apparent decrease in counts in a relatively thin septum (partial-volume effect). Key findings in myocardial perfusion images in the presence of left bundle-branch block that define true positives (ischemia) are reversible perfusion defects (especially at end diastole), a concomitant apical defect, and systolic dysfunction matching the perfusion defect.

Tl-201 myocardial SPECT in differentiation of ischemic from nonischemic dilated cardiomyopathy in patients with left ventricular dysfunction

BACKGROUND

The differentiation between ischemic and nonischemic cardiomyopathy by noninvasive modalities is of clinical importance. Whether thallium 201 single photon emission computed tomography (SPECT) could accurately distinguish the two groups remains unclear.

METHODS AND RESULTS

Twenty-nine patients with chronic heart failure (left ventricular ejection fraction < or =40%), including fourteen patients with ischemic cardiomyopathy and fifteen patients with nonischemic dilated cardiomyopathy, underwent Tl-201 SPECT. The stress protocols included treadmill exercise in 8 patients, dipyridamole in 6 patients, and dobutamine infusion in 15 patients. Myocardial SPECT was interpreted with the use of a 17-segment model and 0- to 4-point scale system. Patients with ischemic cardiomyopathy had higher summed stress defect scores (27.9 +/- 9.4 vs 20.6 +/- 8.9, P =.04), more fixed defect segments (5.9 +/- 2.9 vs 3.8 +/- 2.9, P =.05), and more moderate or severe perfusion defect segments on stress scan (7.2 +/- 2.0 vs 4.5 +/- 2.6, P =.004) than did those with nonischemic dilated cardiomyopathy. However, considerable overlap of the scan patterns between the two groups existed. Moderate or severe perfusion defects on stress scan in at least 7 segments were noted in 71% of patients (10/14) with ischemic cardiomyopathy, as compared with 20% of patients (3/15) with nonischemic cardiomyopathy (P =.016).

CONCLUSIONS

Assessment of Tl-201 myocardial SPECT yields only modest value to distinguish nonischemic dilated cardiomyopathy from ischemic cardiomyopathy in patients with chronic heart failure. This technique cannot clearly differentiate individual patients.