Multicenter evaluation of single-photon emission computed tomography quantification with third-party reconstruction software

Automated identification of uncertain cases in deep learning-based classification of dopamine transporter SPECT to improve clinical utility and acceptance

PURPOSE

Deep convolutional neural networks (CNN) are promising for automatic classification of dopamine transporter (DAT)-SPECT images. Reporting the certainty of CNN-based decisions is highly desired to flag cases that might be misclassified and, therefore, require particularly careful inspection by the user. The aim of the current study was to design and validate a CNN-based system for the identification of uncertain cases.

METHODS

A network ensemble (NE) combining five CNNs was trained for binary classification of [123I]FP-CIT DAT-SPECT images as "normal" or "neurodegeneration-typical reduction" with high accuracy (NE for classification, NEfC). An uncertainty detection module (UDM) was obtained by combining two additional NE, one trained for detection of "reduced" DAT-SPECT with high sensitivity, the other with high specificity. A case was considered "uncertain" if the "high sensitivity" NE and the "high specificity" NE disagreed. An internal "development" dataset of 1740 clinical DAT-SPECT images was used for training (n = 1250) and testing (n = 490). Two independent datasets with different image characteristics were used for testing only (n = 640, 645). Three established approaches for uncertainty detection were used for comparison (sigmoid, dropout, model averaging).

RESULTS

In the test data from the development dataset, the NEfC achieved 98.0% accuracy. 4.3% of all test cases were flagged as "uncertain" by the UDM: 2.5% of the correctly classified cases and 90% of the misclassified cases. NEfC accuracy among "certain" cases was 99.8%. The three comparison methods were less effective in labelling misclassified cases as "uncertain" (40-80%). These findings were confirmed in both additional test datasets.

CONCLUSION

The UDM allows reliable identification of uncertain [123I]FP-CIT SPECT with high risk of misclassification. We recommend that automatic classification of [123I]FP-CIT SPECT images is combined with an UDM to improve clinical utility and acceptance. The proposed UDM method ("high sensitivity versus high specificity") might be useful also for DAT imaging with other ligands and for other binary classification tasks.

99mTc-Labeled FAPI SPECT Imaging in Idiopathic Pulmonary Fibrosis: Preliminary Results

AIM

Idiopathic pulmonary fibrosis (IPF) is associated with a poor prognosis, presenting the most aggressive form of interstitial lung diseases (ILDs). Activated fibroblasts are crucial for pathological processes. Fibroblast activation protein (FAP) inhibitor (FAPI) tracers would be promising imaging agents for these diseases. The purpose of this study was to evaluate a 99mTc-labeled FAPI tracer, 99mTc-HFAPI imaging in IPF patients.

METHODS

Eleven IPF patients (nine males and two females; age range 55-75 year) were included in this pilot study. 99mTc-HFAPI serial whole-body scintigraphy at 5 min, 20 min, 40 min, 1 h, 2 h, 3 h, 4 h, and 6 h was acquired for dynamic biodistribution and dosimetry estimation in seven representative patients. SPECT/CT tomography fusion imaging of the chest region was performed in all patients at 4 h post-injection, which was considered as the optimal acquisition time. Dosimetry was calculated using OLINDA/EXM software (version 2.0; HERMES Medical Solutions). The quantified or semi-quantified standardized uptake values (SUVs) and lesion-to-background ratios (LBRs) of affected lung parenchyma were also calculated. The high-resolution CT (HRCT) stage was determined with visual evaluation, and the total HRCT score of each patient was measured using a weighting factor formula. Pulmonary function tests (PFTs) were recorded as well. Then, the relationships between the 99mTc-HFAPI results, disease extent on HRCT, and PFT results were investigated.

RESULTS

Normal physiological uptake of 99mTc-HFAPI was observed mainly in the liver, intestinal tract, pancreas, gallbladder, and to a lesser extent in the spleen, kidneys, and thyroid, with no apparent retention in the blood circulation at the late time point. The mean injected activity of 99mTc-HFAPI was 813.4 MBq (range 695.6-888.0 MBq). No subjective side effects were noticed. The average whole-body effective dose was 0.0041 mSv/MBq per patient. IPF patients exhibited elevated pulmonary 99mTc-HFAPI uptake in abnormal lung regions, which was correlated with fibrotic regions on HRCT. Among different HRCT stage groups, both SUVmax and LBR showed significant differences (p < 0.001). The higher HRCT stage demonstrated significantly higher SUVmax and LBR. A linear correlation between 99mTc-HFAPI uptake and total HRCT score was observed for SUVmax (r = 0.7839, F = 54.41, p = 0.0094) and LBR (r = 0.7402, F = 56.33, p = 0.0092). 99mTc-HFAPI uptake also had moderate correlations with PFT results.

CONCLUSIONS

Our preliminary data show that the 99mTc-HFAPI SPECT imaging is a promising new imaging modality in IPF patients. Investigations of its clinical value in monitoring disease progression and treatment response are needed in the future.

No impact of attenuation and scatter correction on the interpretation of dopamine transporter SPECT in patients with clinically uncertain parkinsonian syndrome

PURPOSE

The benefit from attenuation and scatter correction (ASC) of dopamine transporter (DAT)-SPECT for the detection of nigrostriatal degeneration in clinical routine is still a matter of debate. The current study evaluated the impact of ASC on visual interpretation and semi-quantitative analysis of DAT-SPECT in a large patient sample.

METHODS

One thousand seven hundred forty consecutive DAT-SPECT with 123I-FP-CIT from clinical routine were included retrospectively. SPECT images were reconstructed iteratively without and with ASC. Attenuation correction was based on uniform attenuation maps, scatter correction on simulation. All SPECT images were categorized with respect to the presence versus the absence of Parkinson-typical reduction of striatal 123I-FP-CIT uptake by three independent readers. Image reading was performed twice to assess intra-reader variability. The specific 123I-FP-CIT binding ratio (SBR) was used for automatic categorization, separately with and without ASC.

RESULTS

The mean proportion of cases with discrepant categorization by the same reader between the two reading sessions was practically the same without and with ASC, about 2.2%. The proportion of DAT-SPECT with discrepant categorization without versus with ASC by the same reader was 1.66% ± 0.50% (1.09-1.95%), not exceeding the benchmark of 2.2% from intra-reader variability. This also applied to automatic categorization of the DAT-SPECT images based on the putamen SBR (1.78% discrepant cases between without versus with ASC).

CONCLUSION

Given the large sample size, the current findings provide strong evidence against a relevant impact of ASC with uniform attenuation and simulation-based scatter correction on the clinical utility of DAT-SPECT to detect nigrostriatal degeneration in patients with clinically uncertain parkinsonian syndrome.

Prognostic biomarkers in prodromal α-synucleinopathies: DAT binding and REM sleep without atonia

BACKGROUND

Isolated rapid eye movement (REM) sleep behaviour disorder (iRBD) is a prodromal state of clinical α-synucleinopathies such as Parkinson's disease and Lewy body dementia. The lead-time until conversion is unknown. The most reliable marker of progression is reduced striatal dopamine transporter (DAT) binding, but low availability of imaging facilities limits general use. Our prospective observational study aimed to relate metrics of REM sleep without atonia (RWA)-a hallmark of RBD-to DAT-binding ratios in a large, homogeneous sample of patients with RBD to explore the utility of RWA as a marker of progression in prodromal α-synucleinopathies.

METHODS

DAT single-photon emission CT (SPECT) and video polysomnography (vPSG) were performed in 221 consecutive patients with clinically suspected RBD.

RESULTS

vPSG confirmed RBD in 176 patients (162 iRBD, 14 phenoconverted, 45 non-synucleinopathies). Specific DAT-binding ratios differed significantly between groups, but showed considerable overlap. Most RWA metrics correlated significantly with DAT-SPECT ratios (eg, Montreal tonic vs most-affected-region: r=-0.525; p<0.001). In patients taking serotonergic/noradrenergic antidepressants or dopaminergic substances or with recent alcohol abuse, correlations were weaker, suggesting a confounding influence, unlike other possible confounders such as beta-blocker use or comorbid sleep apnoea.

CONCLUSIONS

In this large single-centre prospective observational study, we found evidence that DAT-binding ratios in patients with iRBD can be used to describe a continuum in the neurodegenerative process. Overlap with non-synucleinopathies and clinical α-synucleinopathies, however, precludes the use of DAT-binding ratios as a precise diagnostic marker. The parallel course of RWA metrics and DAT-binding ratios suggests in addition to existing data that RWA, part of the routine diagnostic workup in these patients, may represent a marker of progression. Based on our findings, we suggest ranges of RWA values to estimate whether patients are in an early, medium or advanced state within the prodromal phase of α-synucleinopathies, providing them with important information about time until possible conversion.

Multicenter Study of Quantitative SPECT: Reproducibility of 99mTc Quantitation Using a Conjugated-Gradient Minimization Reconstruction Algorithm

This multicenter study aimed to determine the reproducibility of quantitative SPECT images reconstructed using a commercially available method of ordered-subset conjugate-gradient minimization. Methods: A common cylindric phantom containing a 100 kBq/mL concentration of ^99mTc-pertechnetate solution in a volume of 7 L was scanned under standard imaging conditions at 6 institutions using the local clinical protocol of each. Interinstitutional variation among the quantitative SPECT images was evaluated using the coefficient of variation. Dose calibrator accuracy was also investigated by measuring the same lot of commercially available ^99mTc vials at each institution. Results: The respective radioactivity concentrations under standard and clinical conditions ranged from 95.71 ± 0.60 (mean ± SD) to 108.35 ± 0.36 kBq/mL and from 96.78 ± 0.64 to 108.49 ± 0.11 kBq/mL, respectively. Interinstitutional variation in radioactivity concentration was 4.20%. The bias in the radioactivity concentrations in SPECT images was associated with the accuracy of the dose calibrator at each institution. Conclusion: The reproducibility of the commercially available quantitative SPECT reconstruction method is high and comparable to that of PET, for comparatively large (∼7 L), homogeneous objects.

Comparison of commercial dosimetric software platforms in patients treated with 177 Lu-DOTATATE for peptide receptor radionuclide therapy

Purpose

The aim of this study was to quantitatively compare five commercial dosimetric software platforms based on the analysis of clinical datasets of patients who benefited from peptide receptor radionuclide therapy (PRRT) with ¹⁷⁷Lu‐DOTATATE (LUTATHERA^®).

Methods

The dosimetric analysis was performed on two patients during two cycles of PRRT with ¹⁷⁷Lu. Single photon emission computed tomography/computed tomography images were acquired at 4, 24, 72, and 192 h post injection. Reconstructed images were generated using Dosimetry Toolkit^® (DTK) from Xeleris™ and HybridRecon‐Oncology version_1.3_Dicom (HROD) from HERMES. Reconstructed images using DTK were analyzed using the same software to calculate time‐integrated activity coefficients (TIAC), and mean absorbed doses were estimated using OLINDA/EXM V1.0 with mass correction. Reconstructed images from HROD were uploaded into PLANET® OncoDose from DOSIsoft, STRATOS from Phillips, Hybrid Dosimetry Module™ from HERMES, and SurePlan™ MRT from MIM. Organ masses, TIACs, and mean absorbed doses were calculated from each application using their recommendations.

Results

The majority of organ mass estimates varied by <9.5% between all platforms. The highest variability for TIAC results between platforms was seen for the kidneys (28.2%) for the two patients and the two treatment cycles. Relative standard deviations in mean absorbed doses were slightly higher compared with those observed for TIAC, but remained of the same order of magnitude between all platforms.

Conclusions

When applying a similar processing approach, results obtained were of the same order of magnitude regardless of the platforms used. However, the comparison of the performances of currently available platforms is still difficult as they do not all address the same parts of the dosimetric analysis workflow. In addition, the way in which data are handled in each part of the chain from data acquisition to absorbed doses may be different, which complicates the comparison exercise. Therefore, the dissemination of commercial solutions for absorbed dose calculation calls for the development of tools and standards allowing for the comparison of the performances between dosimetric software platforms.

Variability in lutetium-177 SPECT quantification between different state-of-the-art SPECT/CT systems

BACKGROUND

Quantitative SPECT imaging in targeted radionuclide therapy with lutetium-177 holds great potential for individualized treatment based on dose assessment. The establishment of dose-effect relations requires a standardized method for SPECT quantification. The purpose of this multi-center study is to evaluate quantitative accuracy and inter-system variations of different SPECT/CT systems with corresponding commercially available quantitative reconstruction algorithms. This is an important step towards a vendor-independent standard for quantitative lutetium-177 SPECT.

METHODS

Four state-of-the-art SPECT/CT systems were included: Discovery™ NM/CT 670Pro (GE Healthcare), Symbia Intevo™, and two Symbia™ T16 (Siemens Healthineers). Quantitative accuracy and inter-system variations were evaluated by repeatedly scanning a cylindrical phantom with 6 spherical inserts (0.5 - 113 ml). A sphere-to-background activity concentration ratio of 10:1 was used. Acquisition settings were standardized: medium energy collimator, body contour trajectory, photon energy window of 208 keV (± 10%), adjacent 20% lower scatter window, 2 × 64 projections, 128 × 128 matrix size, and 40 s projection time. Reconstructions were performed using GE Evolution with Q.Metrix™, Siemens xSPECT Quant™, Siemens Broad Quantification™ or Siemens Flash3D™ algorithms using vendor recommended settings. In addition, projection data were reconstructed using Hermes SUV SPECT™ with standardized reconstruction settings to obtain a vendor-neutral quantitative reconstruction for all systems. Volumes of interest (VOI) for the spheres were obtained by applying a 50% threshold of the sphere maximum voxel value corrected for background activity. For each sphere, the mean and maximum recovery coefficient (RC_mean and RC_max) of three repeated measurements was calculated, defined as the imaged activity concentration divided by the actual activity concentration. Inter-system variations were defined as the range of RC over all systems.

RESULTS

RC decreased with decreasing sphere volume. Inter-system variations with vendor-specific reconstructions were between 0.06 and 0.41 for RC_mean depending on sphere size (maximum 118% quantification difference), and improved to 0.02-0.19 with vendor-neutral reconstructions (maximum 38% quantification difference).

CONCLUSION

This study shows that eliminating sources of possible variation drastically reduces inter-system variation in quantification. This means that absolute SPECT quantification for ¹⁷⁷Lu is feasible in a multi-center and multi-vendor setting; however, close agreement between vendors and sites is key for multi-center dosimetry and quantitative biomarker studies.

Analysis of quantitative [I-123] mIBG SPECT/CT in a phantom and in patients with neuroblastoma

Purpose

To determine the accuracy of quantitative SPECT, intersystem and interpatient standardized uptake value (SUV) calculation consistency for a manufacturer-independent quantitative SPECT/CT reconstruction algorithm, and the range of SUVs of normal and neoplastic tissue.

Methods

A NEMA body phantom with 6 spheres (ranging 10–37 mm) was filled with a known activity-to-volume ratio and used to determine the contrast recovery coefficient (CRC) for each visible sphere, and the measured SUV accuracy of those spheres and background water solution. One hundred eleven 123I-metaiodobenzylguanidine ([I-123] mIBG) SPECT/CT examinations from 43 patients were reconstructed using SUV SPECT® (HERMES Medical Solutions Inc.); 42 examinations were acquired using a GE Infinia Hawkeye 4 SPECT/CT, and 69 were acquired on a Siemens Symbia Intevo SPECT/CT. Inter scanner SUV analysis of 9 regions of normal [I-123] mIBG tissue uptake was conducted. Intrapatient mean SUV variability was calculated by measuring normal liver uptake within patients scanned on both cameras. The intensity of uptake by neoplastic tissue in the images was quantified using maximum SUV and, if present, compared over time.

Results

The phantom results of the visible spheres and background resulted in accuracy calculations better than 5–10% with CRC correction. Interscanner SUV variability showed no statistical difference (average p value 0.559; range 0.066–1.0) among the 9 normal tissues analyzed. Intrapatient liver mean SUV varied ≤ 16% as calculated for 28 patients (87 examinations) studied on both scanners. In one patient, a thoracic tumor evaluated over 10 time points (18 months) underwent a 74% (3.1/12.0) reduction in maximum SUV with treatment.

Conclusion

The results demonstrate quantitative accuracy to better than 10%, and both consistent SUV calculation between 2 different SPECT/CT scanners for 9 tissues, and low intrapatient measurement variability for quantitative SPECT/CT analysis in a pediatric population with neuroblastoma. Quantitative SPECT/CT offers the opportunity for objective analysis of tumor response using [I-123] mIBG by normalizing the uptake to injected dose and patient weight, as is done for PET.

Towards standardization of absolute SPECT/CT quantification: a multi-center and multi-vendor phantom study

Abstract

Absolute quantification of radiotracer distribution using SPECT/CT imaging is of great importance for dosimetry aimed at personalized radionuclide precision treatment. However, its accuracy depends on many factors. Using phantom measurements, this multi-vendor and multi-center study evaluates the quantitative accuracy and inter-system variability of various SPECT/CT systems as well as the effect of patient size, processing software and reconstruction algorithms on recovery coefficients (RC).

Methods

Five SPECT/CT systems were included: Discovery™ NM/CT 670 Pro (GE Healthcare), Precedence™ 6 (Philips Healthcare), Symbia Intevo™, and Symbia™ T16 (twice) (Siemens Healthineers). Three phantoms were used based on the NEMA IEC body phantom without lung insert simulating body mass indexes (BMI) of 25, 28, and 47 kg/m². Six spheres (0.5–26.5 mL) and background were filled with 0.1 and 0.01 MBq/mL ^99mTc-pertechnetate, respectively. Volumes of interest (VOI) of spheres were obtained by a region growing technique using a 50% threshold of the maximum voxel value corrected for background activity. RC, defined as imaged activity concentration divided by actual activity concentration, were determined for maximum (RC_max) and mean voxel value (RC_mean) in the VOI for each sphere diameter. Inter-system variability was expressed as median absolute deviation (MAD) of RC. Acquisition settings were standardized. Images were reconstructed using vendor-specific 3D iterative reconstruction algorithms with institute-specific settings used in clinical practice and processed using a standardized, in-house developed processing tool based on the SimpleITK framework. Additionally, all data were reconstructed with a vendor-neutral reconstruction algorithm (Hybrid Recon™; Hermes Medical Solutions).

Results

RC decreased with decreasing sphere diameter for each system. Inter-system variability (MAD) was 16 and 17% for RC_mean and RC_max, respectively. Standardized reconstruction decreased this variability to 4 and 5%. High BMI hampers quantification of small lesions (< 10 ml).

Conclusion

Absolute SPECT quantification in a multi-center and multi-vendor setting is feasible, especially when reconstruction protocols are standardized, paving the way for a standard for absolute quantitative SPECT.

Intra- and Interobserver Agreement of SUV SPECT Quantitative SPECT/CT Processing Software, Applied in Clinical Settings for Patients with Solid Renal Tumors

Quantification tools for SPECT/CT are a field of ongoing research among hybrid imaging techniques. A recent multicenter study evaluating SUV SPECT software on phantoms showed that quantitative SPECT/CT is reproducible in Hybrid Viewer software. The aim of our study was to evaluate the intra- and interobserver agreement of quantitative SUV SPECT measurements in clinical settings for patients with solid renal tumors. Methods: The evaluation was part of a study that examined the role of ^99mTc-sestamibi SPECT/CT in the characterization of solid kidney tumors and the differentiation of renal oncocytomas from renal cell carcinomas. Quantitative evaluation of SUV measurements was performed in Hybrid Viewer PDR, version 2.5. Forty-eight renal lesions were identified and examined twice by 2 independent readers. The agreement of the SUV measurements, concerning renal tumors and adjacent renal parenchyma, was evaluated using the intraclass correlation coefficient (ICC). Results: ICC for SUV_max measurements by the same reader was 97%-99% for solid renal tumors and 92%-98% for ipsilateral healthy renal parenchyma. ICC for SUV_max measurements between readers was 87%-89% for solid renal tumors and 72%-73% for ipsilateral healthy renal parenchyma. Estimated ICC for SUV_mean measurements of solid renal tumors was 95%-98% for the same reader and 86%-89% between readers. Similar results were found for SUV_peak measurements. Conclusion: The high ICCs indicate a strong agreement among SUV measurements for patients with solid renal lesions undergoing ^99mTc-sestamibi SPECT/CT, not only for measurements by the same reader but also for measurements between 2 different readers.

Spatial dependence of activity concentration recovery for a conjugate gradient (Siemens xSPECT) algorithm using manufacturer-defined reconstruction presets

Siemens absolute quantitative reconstruction, xSPECT, is available with manufacturer-defined reconstruction presets to assist with optimization. This phantom study evaluates the impact of these presets on the spatial dependence of activity concentration recovery (ACR). Single-photon emission computed tomography/computed tomography scans of a 5 : 1 and 10 : 1 (sphere : background) contrast NEMA phantom were performed on a Siemens Intevo 6. Three sphere position configurations, achieved by rotating the sphere mount through 0°, 120° and 240°, were used and three replicate images of each configuration were acquired. xSPECT reconstruction was performed using 'Fast', 'Standard' and 'Best' presets. Maximum voxel and A50 threshold ACR were measured in each sphere. The average ACR per sphere was calculated across replicates. Percentage variation of ACR, about this average, for each sphere within a given configuration across replicates and also alternative configurations was calculated. Within a given sphere configuration, percentage variation for maximum voxel ACR in like-for-like spheres across replicates was within 11% for all three presets across all sphere sizes, and within 3% for 10 : 1 and 9% for 5 : 1 contrast in the three largest spheres. Substantial variation of ACR was observed when comparing like-for-like spheres in different configurations. In the three largest spheres, variation in maximum ACR of up to 35 and 32% was measured for 10 : 1 and 5 : 1 contrast, respectively. Variation in activity concentration may be substantially greater than perceived from using a single phantom configuration. The spatial dependence observed using the manufacturer presets highlights the need for evaluation of user-defined reconstruction parameters.

Evaluation of quantitative 123I and 131I SPECT with Monte Carlo-based down-scatter compensation

OBJECTIVE

Quantitative I and I single-photon emission computed tomography (SPECT) is hampered by down-scatter from the high-energy peaks. This paper presents a down-scatter compensation method, where down-scatter generated in the patient and gamma camera collimator and detector is modelled using Monte Carlo simulation in the ordered subsets expectation maximization SPECT reconstruction algorithm.

MATERIALS AND METHODS

The new down-scatter compensation method was compared with conventional triple energy window (TEW) scatter compensation and Gaussian convolution-based forced detection Monte Carlo methods. The comparison was made with the NEMA-IEC phantom using six spherical inserts (diameters from 10 to 37 mm) and a lung compartment. The phantom was filled with I and I solutions to known sphere-to-background concentration ratios. Spherical volumes of interest with the same diameter as the inserts were drawn on the images, and recovery coefficients for the spheres were calculated in addition to lung-to-background ratio.

RESULTS

The new down-scatter compensation method provided higher recovery coefficients than the TEW scatter compensation or Gaussian convolution-based forced detection Monte Carlo algorithm for both isotopes. Background activity concentration could be accurately estimated with the new down-scatter compensation method and with the TEW scatter compensation, whereas activity concentration of the spheres was severely underestimated even with the new method.

CONCLUSION

Down-scatter compensation with Monte Carlo-simulation effectively reduces down-scatter effects in I and I SPECT imaging.

Implementation of GPU accelerated SPECT reconstruction with Monte Carlo-based scatter correction

OBJECTIVE

Statistical SPECT reconstruction can be very time-consuming especially when compensations for collimator and detector response, attenuation, and scatter are included in the reconstruction. This work proposes an accelerated SPECT reconstruction algorithm based on graphics processing unit (GPU) processing.

METHODS

Ordered subset expectation maximization (OSEM) algorithm with CT-based attenuation modelling, depth-dependent Gaussian convolution-based collimator-detector response modelling, and Monte Carlo-based scatter compensation was implemented using OpenCL. The OpenCL implementation was compared against the existing multi-threaded OSEM implementation running on a central processing unit (CPU) in terms of scatter-to-primary ratios, standardized uptake values (SUVs), and processing speed using mathematical phantoms and clinical multi-bed bone SPECT/CT studies.

RESULTS

The difference in scatter-to-primary ratios, visual appearance, and SUVs between GPU and CPU implementations was minor. On the other hand, at its best, the GPU implementation was noticed to be 24 times faster than the multi-threaded CPU version on a normal 128 × 128 matrix size 3 bed bone SPECT/CT data set when compensations for collimator and detector response, attenuation, and scatter were included.

CONCLUSIONS

GPU SPECT reconstructions show great promise as an every day clinical reconstruction tool.

Phantom and clinical evaluation of the effect of full Monte Carlo collimator modelling in post-SIRT yttrium-90 Bremsstrahlung SPECT imaging

Background

Post-therapy SPECT/CT imaging of ⁹⁰Y microspheres delivered to hepatic malignancies is difficult, owing to the continuous, high-energy Bremsstrahlung spectrum emitted by ⁹⁰Y. This study aimed to evaluate the utility of a commercially available software package (HybridRecon, Hermes Medical Solutions AB) which incorporates full Monte Carlo collimator modelling. Analysis of image quality was performed on both phantom and clinical images in order to ultimately provide a recommendation of an optimum reconstruction for post-therapy ⁹⁰Y microsphere SPECT/CT imaging.

A 3D-printed anthropomorphic liver phantom was filled with ⁹⁰Y with a sphere-to-background ratio of 4:1 and imaged on a GE Discovery 670 SPECT/CT camera. Datasets were reconstructed using ordered-subsets expectation maximization (OSEM) 1–7 iterations in order to identify the optimal OSEM reconstruction (5 iterations, 15 subsets). Quantitative analysis was subsequently carried out on phantom datasets obtained using four reconstruction algorithms: the default OSEM protocol (2 iterations, 10 subsets) and the optimised OSEM protocol, both with and without full Monte Carlo collimator modelling. The quantitative metrics contrast recovery (CR) and background variability (BV) were calculated.

The four algorithms were then used to retrospectively reconstruct 10 selective internal radiation therapy (SIRT) patient datasets which were subsequently blind scored for image quality by a consultant radiologist.

Results

The optimised OSEM reconstruction (5 iterations, 15 subsets with full MC collimator modelling) increased the CR by 42% (p < 0.001) compared to the default OSEM protocol (2 iterations, 10 subsets). The use of full Monte Carlo collimator modelling was shown to further improve CR by 14% (30 mm sphere, CR = 90%, p < 0.05).

The consultant radiologist had a significant preference for the optimised OSEM over the default OSEM protocol (p < 0.001), with the optimised OSEM being the favoured reconstruction in every one of the 10 clinical cases presented.

Conclusions

OSEM (5 iterations, 15 subsets) with full Monte Carlo collimator modelling is quantitatively the optimal image reconstruction for post-SIRT ⁹⁰Y Bremsstrahlung SPECT/CT imaging. The use of full Monte Carlo collimator modelling for correction of image-degrading effects significantly increases contrast recovery without degrading clinical image quality.