The impact of system matrix dimension on small FOV SPECT reconstruction with truncated projections

DuDoCFNet: Dual-Domain Coarse-to-Fine Progressive Network for Simultaneous Denoising, Limited-View Reconstruction, and Attenuation Correction of Cardiac SPECT

Single-Photon Emission Computed Tomography (SPECT) is widely applied for the diagnosis of coronary artery diseases. Low-dose (LD) SPECT aims to minimize radiation exposure but leads to increased image noise. Limited-view (LV) SPECT, such as the latest GE MyoSPECT ES system, enables accelerated scanning and reduces hardware expenses but degrades reconstruction accuracy. Additionally, Computed Tomography (CT) is commonly used to derive attenuation maps ( μ -maps) for attenuation correction (AC) of cardiac SPECT, but it will introduce additional radiation exposure and SPECT-CT misalignments. Although various methods have been developed to solely focus on LD denoising, LV reconstruction, or CT-free AC in SPECT, the solution for simultaneously addressing these tasks remains challenging and under-explored. Furthermore, it is essential to explore the potential of fusing cross-domain and cross-modality information across these interrelated tasks to further enhance the accuracy of each task. Thus, we propose a Dual-Domain Coarse-to-Fine Progressive Network (DuDoCFNet), a multi-task learning method for simultaneous LD denoising, LV reconstruction, and CT-free μ -map generation of cardiac SPECT. Paired dual-domain networks in DuDoCFNet are cascaded using a multi-layer fusion mechanism for cross-domain and cross-modality feature fusion. Two-stage progressive learning strategies are applied in both projection and image domains to achieve coarse-to-fine estimations of SPECT projections and CT-derived μ -maps. Our experiments demonstrate DuDoCFNet's superior accuracy in estimating projections, generating μ -maps, and AC reconstructions compared to existing single- or multi-task learning methods, under various iterations and LD levels. The source code of this work is available at https://github.com/XiongchaoChen/DuDoCFNet-MultiTask.

Direct and indirect strategies of deep-learning-based attenuation correction for general purpose and dedicated cardiac SPECT

PURPOSE

Deep-learning-based attenuation correction (AC) for SPECT includes both indirect and direct approaches. Indirect approaches generate attenuation maps (μ-maps) from emission images, while direct approaches predict AC images directly from non-attenuation-corrected (NAC) images without μ-maps. For dedicated cardiac SPECT scanners with CZT detectors, indirect approaches are challenging due to the limited field-of-view (FOV). In this work, we aim to 1) first develop novel indirect approaches to improve the AC performance for dedicated SPECT; and 2) compare the AC performance between direct and indirect approaches for both general purpose and dedicated SPECT.

METHODS

For dedicated SPECT, we developed strategies to predict truncated μ-maps from NAC images reconstructed with a small matrix, or full μ-maps from NAC images reconstructed with a large matrix using 270 anonymized clinical studies scanned on a GE Discovery NM/CT 570c SPECT/CT. For general purpose SPECT, we implemented direct and indirect approaches using 400 anonymized clinical studies scanned on a GE NM/CT 850c SPECT/CT. NAC images in both photopeak and scatter windows were input to predict μ-maps or AC images.

RESULTS

For dedicated SPECT, the averaged normalized mean square error (NMSE) using our proposed strategies with full μ-maps was 1.20 ± 0.72% as compared to 2.21 ± 1.17% using the previous direct approaches. The polar map absolute percent error (APE) using our approaches was 3.24 ± 2.79% (R² = 0.9499) as compared to 4.77 ± 3.96% (R² = 0.9213) using direct approaches. For general purpose SPECT, the averaged NMSE of the predicted AC images using the direct approaches was 2.57 ± 1.06% as compared to 1.37 ± 1.16% using the indirect approaches.

CONCLUSIONS

We developed strategies of generating μ-maps for dedicated cardiac SPECT with small FOV. For both general purpose and dedicated SPECT, indirect approaches showed superior performance of AC than direct approaches.

SPECTnet: a deep learning neural network for SPECT image reconstruction

Background

Single photon emission computed tomography (SPECT) is an important functional tool for clinical diagnosis and scientific research of brain disorders, but suffers from limited spatial resolution and high noise due to hardware design and imaging physics. The present study is to develop a deep learning technique for SPECT image reconstruction that directly converts raw projection data to image with high resolution and low noise, while an efficient training method specifically applicable to medical image reconstruction is presented.

Methods

A custom software was developed to generate 20,000 2-D brain phantoms, of which 16,000 were used to train the neural network, 2,000 for validation, and the final 2,000 for testing. To reduce development difficulty, a two-step training strategy for network design was adopted. We first compressed full-size activity image (128×128 pixels) to a one-D vector consisting of 256×1 pixels, accomplished by an autoencoder (AE) consisting of an encoder and a decoder. The vector is a good representation of the full-size image in a lower-dimensional space and was used as a compact label to develop the second network that maps between the projection-data domain and the vector domain. Since the label had 256 pixels only, the second network was compact and easy to converge. The second network, when successfully developed, was connected to the decoder (a portion of AE) to decompress the vector to a regular 128×128 image. Therefore, a complex network was essentially divided into two compact neural networks trained separately in sequence but eventually connectable.

Results

A total of 2,000 test examples, a synthetic brain phantom, and de-identified patient data were used to validate SPECTnet. Results obtained from SPECTnet were compared with those obtained from our clinic OS-EM method. Images with lower noise and more accurate information in the uptake areas were obtained by SPECTnet.

Conclusions

The challenge of developing a complex deep neural network is reduced by training two separate compact connectable networks. The combination of the two networks forms the full version of SPECTnet. Results show that the developed neural network can produce more accurate SPECT images.

Patient arm position during quantitative bone single-photon emission computed tomography/computed tomography acquisition can affect image quality and quantitative accuracy: a phantom study

PURPOSE

The present study used a phantom to determine the effects of various arm positions on bone SPECT/computed tomography (CT) images and the optimal arm position to acquire good-quality and quantitatively accurate images.

MATERIALS AND METHODS

We designed a phantom study of five simulated arm positions that are assumed during SPECT image acquisition. All SPECT data were acquired during a total of 120 projections of 10 and 100 s/view over 360° in a non-circular mode and reconstructed using Flash 3D (Siemens Healthineers). We evaluated contrast (QH,17 mm), image noise (NB,17 mm), contrast-to-noise ratios (QNRs), and visual scores according to the guidelines for bone SPECT acquisition protocols published by the Japanese Society of Nuclear Medicine Technology. The SUVmean, SUVmax, and SUVpeak were calculated and quantitative errors were evaluated using the recovery coefficient (RC) and the root means square error (RMSE).

RESULTS

The spatial resolution of SPECT images was better when the arms were down than raised with simulated shoulder disorders. Raised arms with shoulder disorders significantly increased the NB,17 mm and decreased the QH,17 mm, and the QNR in each image differed over a range from 2.2 to 5.2. The visual score was >1.5 with the arms down, raised normally, and raised with moderate shoulder disorders. The SUVmax and SUVpeak were overestimated compared with 100-min data for all images, whereas SUVmean was underestimated. Raised arms with a shoulder disorder decreased RCmax, and RCmean and RCpeak suppressed differences among arm positions. In addition, RMSE with the arms down and raised normally were close to that for 100-min data.

CONCLUSION

Bone SPECT images with good quality and quantitative accuracy can be acquired with patients holding their arms down by their sides. This will help patients with shoulder pain who have difficulties raising their arms.

Best practice for the nuclear medicine technologist in CT-based attenuation correction and calcium score for nuclear cardiology

The use of hybrid systems is increasingly growing in Europe and this is progressively important for the final result of diagnostic tests. As an integral part of the hybrid imaging system, computed tomography (CT) plays a crucial role in myocardial perfusion imaging diagnostics. Throughout Europe, a variety of equipment is available and also different university curricula of the nuclear medicine technologist are observed. Hence, the Technologist Committee of the European Association of Nuclear Medicine proposes to identify, through a bibliographic review, the recommendations for best practice in computed tomography applied to attenuation correction and calcium score in myocardial perfusion imaging, which courses in the set of knowledge, skills, and competencies for nuclear medicine technologists. This document aims at providing recommendations for CT acquisition protocols and CT image optimization in nuclear cardiology.

Direct List Mode Parametric Reconstruction for Dynamic Cardiac SPECT

Recently introduced stationary dedicated cardiac SPECT scanners provide new opportunities to quantify myocardial blood flow (MBF) using dynamic SPECT. However, comparing to PET, the low sensitivity of SPECT scanners affects MBF quantification due to the high noise level, especially for 201 Thallium (201Tl) due to its typically low injected dose. The conventional indirect method for generating parametric images typically starts by reconstructing a time series of frame images followed by fitting the time-activity curve (TAC) for each voxel or segment with an appropriate kinetic model. The indirect method is simple and easy to implement; however, it usually suffers from substantial image noise that could also lead to bias. In this paper, we developed a list mode direct parametric image reconstruction algorithm to substantially reduce noise in MBF quantification using dynamic SPECT and allow for patient radiation dose reduction. GPU-based parallel computing was used to achieve more than 2000-fold acceleration. The proposed method was evaluated in both simulation and in vivo canine studies. Compared with the indirect method, the proposed direct method achieved substantially lower image noise and variability, particularly at large number of iterations and at low-count levels.

Recent advances in cardiac SPECT instrumentation and imaging methods

Cardiac SPECT continues to play a critical role in detecting and managing cardiovascular disease, in particularly coronary artery disease (CAD) (Jaarsma et al 2012 J. Am. Coll. Cardiol. 59 1719-28), (Agostini et al 2016 Eur. J. Nucl. Med. Mol. Imaging 43 2423-32). While conventional dual-head SPECT scanners using parallel-hole collimators and scintillation crystals with photomultiplier tubes are still the workhorse of cardiac SPECT, they have the limitations of low photon sensitivity (~130 count s^-1 MBq^-1), poor image resolution (~15 mm) (Imbert et al 2012 J. Nucl. Med. 53 1897-903), relatively long acquisition time, inefficient use of the detector, high radiation dose, etc. Recently our field observed an exciting growth of new developments of dedicated cardiac scanners and collimators, as well as novel imaging algorithms for quantitative cardiac SPECT. These developments have opened doors to new applications with potential clinical impact, including ultra-low-dose imaging, absolute quantification of myocardial blood flow (MBF) and coronary flow reserve (CFR), multi-radionuclide imaging, and improved image quality as a result of attenuation, scatter, motion, and partial volume corrections (PVCs). In this article, we review the recent advances in cardiac SPECT instrumentation and imaging methods. This review mainly focuses on the most recent developments published since 2012 and points to the future of cardiac SPECT from an imaging physics perspective.

Quantification of intramyocardial blood volume with 99mTc-RBC SPECT-CT imaging: A preclinical study

BACKGROUND

Currently, there is no established non-invasive imaging approach to directly evaluate myocardial microcirculatory function in order to diagnose microvascular disease independent of co-existing epicardial disease. In this work, we developed a methodological framework for quantification of intramyocardial blood volume (IMBV) as a novel index of microcirculatory function with SPECT/CT imaging of 99mTc-labeled red blood cells (RBCs).

METHODS

Dual-gated myocardial SPECT/CT equilibrium imaging of 99mTc-RBCs was performed on twelve canines under resting conditions. Five correction schemes were studied: cardiac gating with no other corrections (CG), CG with attenuation correction (CG + AC), CG + AC with scatter correction (CG + AC + SC), dual cardiorespiratory gating with AC + SC (DG + AC + SC), and DG + AC + SC with partial volume correction (DG + AC + SC + PVC). Quantification of IMBV using each approach was evaluated in comparison to those obtained from all corrections. The in vivo SPECT estimates of IMBV values were validated against those obtained from ex vivo microCT imaging of the casted hearts.

RESULTS

The estimated IMBV with all corrections was 0.15 ± 0.03 for the end-diastolic phase and 0.11 ± 0.03 for the end-systolic phase. The cycle-dependent change in IMBV (ΔIMBV) with all corrections was 23.9 ± 8.6%. Schemes that applied no correction or partial correction resulted in significant over-estimation of IMBV and significant under-underestimation of ΔIMBV. Estimates of IMBV and ΔIMBV using all corrections were consistent with values reported in the literature using invasive techniques. In vivo SPECT estimates of IMBV strongly correlated (R2 ≥ 0.70) with ex vivo measures for the various correction schemes, while the fully corrected scheme yielded the smallest bias.

CONCLUSIONS

Non-invasive quantification of IMBV is feasible using 99mTc-RBCs SPECT/CT imaging, however, requires full compensation of physical degradation factors.

Performance analysis of a high-sensitivity multi-pinhole cardiac SPECT system with hemi-ellipsoid detectors

PURPOSE

Single-photon emission computed tomography (SPECT) is a noninvasive imaging modality, used in myocardial perfusion imaging. The challenges facing the majority of clinical SPECT systems are low sensitivity, poor resolution, and the relatively high radiation dose to the patient. New generation systems (GE Discovery, DSPECT) dedicated to cardiac imaging improve sensitivity by a factor of 5-8. This improvement can be used to decrease acquisition time and/or dose. However, in the case of ultra-low dose (~3 mCi) injections, acquisition times are still significantly long, taking 10-12 min. The purpose of this work is to investigate a new gamma camera design with 21 hemi-ellipsoid detectors each with a pinhole collimator for cardiac SPECT for further improvement in sensitivity and resolution and reduced patient exposures and imaging times.

METHODS

To evaluate the resolution of our hemi-ellipsoid system, GATE Monte-Carlo simulations were performed on point-sources, rod-sources, and NCAT phantoms. For average full-width-half-maximum (FWHM) equivalence with base flat-detector, the pinhole-diameter for the curved hemi-ellipsoid detector was found to be 8.68 mm, an operating pinhole-diameter nominally expected to be ~3 times more sensitive than state-of-the-art systems. Rod-sources equally spaced within the region of interest were acquired with a 21-detector system and reconstructed with our multi-pinhole (MPH) iterative OSEM algorithm with collimator resolution recovery. The results were compared with the results of a state-of-the-art system (GE Discovery) available in the literature. The system was also evaluated using the mathematical anthropomorphic NCAT (NURBS-based Cardiac Torso; Segars et al. IEEE Trans Nucl Sci. 1999;46:503-506) phantom with a full (clinical)-dose acquisition (25 mCi) for 2 min and an ultra-low dose acquisition of 3 mCi for 5.44 min. The estimated left ventricle (LV) counts were compared with the available literature on a state-of-the-art system (DSPECT). FWHM of the LV wall on MPH-OSEM-reconstructed images with collimator resolution recovery was estimated.

RESULTS

On acquired rod-sources, the average resolution (FWHM) after reconstruction with resolution recovery in the entire region of interest (ROI) for cardiac imaging was on the average 4.44 mm (±2.84), compared to 6.9 mm (±1 mm) reported for GE Discovery (Kennedy et al., J Nucl Cardiol. 2014:21:443-452). For NCAT studies, improved sensitivity allowed a full-dose (25 mCi) 2-min acquisition (Ell8.68mmFD) which yielded 3.79 M LV counts. This is ~3.35 times higher compared to 1.13 M LV counts acquired in 2 min for clinical full dose for state-of-the-art DSPECT. The increased sensitivity also allowed an ultra-low dose acquisition protocol (Ell8.68 mmULD), 3 mCi (eight times less injected dose) in 5.44 min. This ultra-low dose protocol yielded ~1.23 M LV counts which was comparable to the full-dose 2-min acquisition for DSPECT. The estimated NCAT average FWHM at the LV wall after 12 iterations of the OSEM reconstruction was 4.95 and 5.66 mm around the mid-short-axis slices for Ell8.68mmFD and Ell8.68mmULD, respectively.

CONCLUSION

Our Monte-Carlo simulation studies and reconstruction suggest using (inverted wineglass sized) hemi-ellipsoid detectors with pinhole collimators can increase the sensitivity ~3.35 times over the new generation of dedicated cardiac SPECT systems, while also improving the reconstructed resolution for rod-sources with an average of 4.44 mm in region of interest. The extra sensitivity may be used for ultra-low dose imaging (3 mCi) at ~5.44 min for comparable clinical counts as state-of-the-art systems.

Application of the 4-D XCAT Phantoms in Biomedical Imaging and Beyond

The four-dimensional (4-D) eXtended CArdiac-Torso (XCAT) series of phantoms was developed to provide accurate computerized models of the human anatomy and physiology. The XCAT series encompasses a vast population of phantoms of varying ages from newborn to adult, each including parameterized models for the cardiac and respiratory motions. With great flexibility in the XCAT's design, any number of body sizes, different anatomies, cardiac or respiratory motions or patterns, patient positions and orientations, and spatial resolutions can be simulated. As such, the XCAT phantoms are gaining a wide use in biomedical imaging research. There they can provide a virtual patient base from which to quantitatively evaluate and improve imaging instrumentation, data acquisition, techniques, and image reconstruction and processing methods which can lead to improved image quality and more accurate clinical diagnoses. The phantoms have also found great use in radiation dosimetry, radiation therapy, medical device design, and even the security and defense industry. This review paper highlights some specific areas in which the XCAT phantoms have found use within biomedical imaging and other fields. From these examples, we illustrate the increasingly important role that computerized phantoms and computer simulation are playing in the research community.

Noise suppressed partial volume correction for cardiac SPECT/CT

PURPOSE

Partial volume correction (PVC) methods typically improve quantification at the expense of increased image noise and reduced reproducibility. In this study, the authors developed a novel voxel-based PVC method that incorporates anatomical knowledge to improve quantification while suppressing noise for cardiac SPECT/CT imaging.

METHODS

In the proposed method, the SPECT images were first reconstructed using anatomical-based maximum a posteriori (AMAP) with Bowsher's prior to penalize noise while preserving boundaries. A sequential voxel-by-voxel PVC approach (Yang's method) was then applied on the AMAP reconstruction using a template response. This template response was obtained by forward projecting a template derived from a contrast-enhanced CT image, and then reconstructed using AMAP to model the partial volume effects (PVEs) introduced by both the system resolution and the smoothing applied during reconstruction. To evaluate the proposed noise suppressed PVC (NS-PVC), the authors first simulated two types of cardiac SPECT studies: a (99m)Tc-tetrofosmin myocardial perfusion scan and a (99m)Tc-labeled red blood cell (RBC) scan on a dedicated cardiac multiple pinhole SPECT/CT at both high and low count levels. The authors then applied the proposed method on a canine equilibrium blood pool study following injection with (99m)Tc-RBCs at different count levels by rebinning the list-mode data into shorter acquisitions. The proposed method was compared to MLEM reconstruction without PVC, two conventional PVC methods, including Yang's method and multitarget correction (MTC) applied on the MLEM reconstruction, and AMAP reconstruction without PVC.

RESULTS

The results showed that the Yang's method improved quantification, however, yielded increased noise and reduced reproducibility in the regions with higher activity. MTC corrected for PVE on high count data with amplified noise, although yielded the worst performance among all the methods tested on low-count data. AMAP effectively suppressed noise and reduced the spill-in effect in the low activity regions. However it was unable to reduce the spill-out effect in high activity regions. NS-PVC yielded superior performance in terms of both quantitative assessment and visual image quality while improving reproducibility.

CONCLUSIONS

The results suggest that NS-PVC may be a promising PVC algorithm for application in low-dose protocols, and in gated and dynamic cardiac studies with low counts.

Quantitative Analysis of Dynamic 123I-mIBG SPECT Imaging Data in Healthy Humans with a Population-Based Metabolite Correction Method

Conventional 2-dimensional planar imaging of (123)I-metaiodobenzylguanidine ((123)I-mIBG) is not fully quantitative. To develop a more accurate quantitative imaging approach, we investigated dynamic SPECT imaging with kinetic modeling in healthy humans to obtain the myocardial volume of distribution (VT) for (123)I-mIBG.

METHODS

Twelve healthy humans underwent 5 serial 15-min SPECT scans at 0, 15, 90, 120, and 180 min after bolus injection of (123)I-mIBG on a hybrid cadmium zinc telluride SPECT/CT system. Serial venous blood samples were obtained for radioactivity measurement and radiometabolite analysis. List-mode data of all the scans were binned into frames and reconstructed with attenuation and scatter corrections. Myocardial and blood-pool volumes of interest were drawn on the reconstructed images to derive the myocardial time-activity curve and input function. A population-based blood-to-plasma ratio (BPR) curve was generated. Both the population-based metabolite correction (PBMC) and the individual metabolite correction (IMC) curves were generated for comparison. VT values were obtained from different compartment models, using different input functions with and without metabolite and BPR corrections.

RESULTS

The BPR curve reached the peak value of 2.1 at 13 min after injection. Parent fraction was approximately 58% ± 13% at 15 min and stabilized at approximately 40% ± 5% by 180 min after injection. Two radiometabolite species were observed. When the reversible 2-tissue-compartment fit was used, the mean VT value was 29.0 ± 12.4 mL/cm(3) with BPR correction and PBMC, a 188% ± 32% increase compared with that without corrections. There was significant difference in VT with BPR correction (P = 2.3e-04) as well as with PBMC (P = 1.6e-05). The mean difference in VT between PBMC and IMC was -3% ± 8%, which was insignificant (P = 0.39). The intersubject coefficients of variation after PBMC (43%) and IMC (42%) were similar.

CONCLUSION

The myocardial VT of (123)I-mIBG was established in healthy humans for the first time. Accurate kinetic modeling of (123)I-mIBG requires both BPR and metabolite corrections. Population-based BPR correction and metabolite correction curves were developed, allowing more convenient absolute quantification of dynamic (123)I-mIBG SPECT images.