CT-free attenuation correction of 13N-ammonia cardiac PET images using conditional denoising diffusion implicit model with logarithmic linear normalization.

An earlier study demonstrated comparable lesion detection between attenuation-corrected (AC) and nonattenuation-corrected (NAC) 18F-sodium fluoride (NaF) PET images, which is relevant for computed tomography (CT) radiation dose-saving. However, this finding may not be applicable to newer systems. The aim was to compare lesion detection between AC and NAC NaF PET images on modern PET-CT systems. One expert and one nonexpert observer retrospectively surveyed NaF PET data in 25 breast cancer patients. At both lesion and patient level, each observer classified bone abnormalities as malignant, equivocal or benign, from NAC and AC PET images in the absence of CT. Expert interpretation of NaF PET-CT with the review of all diagnostic imaging/pathology reports for at least the subsequent 12 months provided reference standard metastases status at the patient level. Two-tailed Wilcoxon signed-rank tests measured statistically significant differences in total lesion detection between AC and NAC PET. Quadratic-weighted kappa score measured agreement in patient metastases status between observers. On a lesion-basis, AC PET images showed significantly more lesions than NAC for both the expert (122 versus 96; P = 0.002) and nonexpert (146 versus 132; P = 0.036) observers, with a large number of patients demonstrating disparity between AC and NAC images. For metastases status at the patient level without CT, NAC PET showed slightly better diagnostic accuracy than AC due to fewer false-positive results, as fewer lesions were identified. AC PET data provided superior lesion detection to NAC in NaF bone examinations and are thus required for clinical interpretation.

The aim of the present study was to evaluate whether computed tomography based-attenuation correction (CT-AC) provides any advantage over non-attenuation-corrected (NAC) images for qualitative and quantitative analysis of single photon emission tomography (SPECT) myocardial perfusion imaging (MPI). We retrospectively evaluated data of 171 patients who underwent stress rest MPI SPECT/CT as per standard protocol. Angiography done within ±3 months of MPI was taken as reference standard. Two readers independently evaluated CT-AC and NAC images. Receiver operating characteristic curve analysis was done using ≥50% and ≥70% stenosis as cutoff. The size and severity of perfusion defects were also compared on CT-AC and NAC images. For both readers, the area under the receiver operating characteristic curve was larger for CT-AC images than for NAC images at both ≥50% and ≥70% cutoff, but the difference was not significant. CT-AC images had significantly lower sensitivity for detecting right coronary artery disease compared with NAC (29% vs. 50% for reader 1 and 25.8% vs. 43.2% for reader 2). However, the specificity improved with CT-AC. Inferior defects were significantly smaller in CT-AC than NAC (P = 0.0002), with no significant difference for anterior defects (P = 0.544). There was significant variation in severity between CT-AC and NAC images for both overall (P = 0.001) as well as for inferior defects (P = 0.0007), but not for anterior defects (P = 0.279). In our study, the CT-based AC improved the specificity but decreased the sensitivity leading to nonsignificant improvement in overall diagnostic accuracy of Tc-99m sestamibi/tetrofosmin MPI.

Heart disease is a leading cause of mortality. Currently, PET is the gold standard for determining the viability of heart tissue. CT is gaining popularity as a transmission (Tx) source for attenuation correction (AC), but due to the differences in scan duration between the emission and Tx scans, breathing motion can cause mis-registration leading to errors in the final image. We explored this problem in the context of a large animal model chosen for its similarities to human physiology. This study investigates the effects of CT based AC (CTAC) and the use of 4D CT for respiratory gated AC, in an in vivo canine model of cardiac PET. Five 22 kg dogs were sedated, ventilated, and injected with 250 MBq of FDG. 4D CT Tx, gated PET, and /sup 68/Ge Tx (GeAC) scans were acquired. Both the 4D CT and the gated PET scans were broken into 15 frames of 0.5 seconds in length. The data were reconstructed using OSEM, applying five different AC maps: Ge, end expiration CT, end inspiration CT, average CT, and phase-matched CT. Ungated PET images were then created by summing the reconstructed gates together. Tendencies were present for end-expiration CTAC to cause over corrections on the anterior side of the heart, while the same was true for end-inspiration CTAC on the inferior side of the heart. On average we found regional differences of up to 3%, though in individual cases these differences were up to 20% on the anterior side and -10% on the inferior side of the heart. Comparing the average CTAC to the phase-matched CT corrected image showed both data sets to be nearly identical. Finally, when the average CTAC was compared to the GeAC image, differences were found along the lateral side of the heart. On average these percentage differences were up to 4%, however in individual cases they were up to 20%. Overall the observed tendencies indicate that there are respiratory motion artifacts, which can be large in individual cases.

Access to cardiac PET/CT by sarcoidosis patients and cost-effectiveness analysis of cardiac PET/MR compared to the standard of care

Auto-segmentation of pelvic organs at risk on 0.35T MRI using 2D and 3D Generative Adversarial Network models

Cardiac PET/CT imaging is often performed in patients with pacemakers and implantable cardioverter defibrillator (ICD) leads. However, metallic implants usually produce artefacts on CT images which might propagate to CT-based attenuation-corrected (CTAC) PET images. The impact of metal artefact reduction (MAR) for CTAC of cardiac PET/CT images in the presence of pacemaker, ICD and ECG leads was investigated using both qualitative and quantitative analysis in phantom and clinical studies. The study included 14 patients with various leads undergoing perfusion and viability examinations using dedicated cardiac PET/CT protocols. The PET data were corrected for attenuation using both artefactual CT images and CT images corrected using the MAR algorithm. The severity and magnitude of metallic artefacts arising from these leads were assessed on both linear attenuation coefficient maps (μ-maps) and attenuation-corrected PET images. CT and PET emission data were obtained using an anthropomorphic thorax phantom and a dedicated heart phantom made in-house incorporating pacemaker and ICD leads attached at the right ventricle of the heart. Volume of interest-based analysis and regression plots were performed for regions related to the lead locations. Bull's eye view analysis was also performed on PET images corrected for attenuation with and without the MAR algorithm. In clinical studies, the visual assessment of PET images by experienced physicians and quantitative analysis did not reveal erroneous interpretation of the tracer distribution or significant differences when PET images were corrected for attenuation with and without MAR. In phantom studies, the mean differences between tracer uptake obtained without and with MAR were 10.16±2.1% and 6.86±2.1% in the segments of the heart in the vicinity of metallic ICD or pacemaker leads, and were 4.43±0.5% and 2.98±0.5% in segments far from the leads. Although the MAR algorithm was able to effectively improve the quality of μ-maps, its clinical impact on the interpretation of PET images was not significant. Therefore cardiac PET images corrected for attenuation using CTAC in the presence of metallic leads can be interpreted without correction for metal artefacts. It should however be emphasized that in some special cases with multiple ICD leads attached to the myocardium wall, MAR might be useful for accurate attenuation correction.

The goal of this work is to demonstrate the feasibility of directly generating attenuation-corrected PET images from non-attenuation-corrected (NAC) PET images for both rest and stress-state static or dynamic [13N]ammonia MP PET based on a generative adversarial network. We recruited 60 subjects for rest-only scans and 14 subjects for rest-stress scans, all of whom underwent [13N]ammonia cardiac PET/CT examinations to acquire static and dynamic frames with both 3D NAC and CT-based AC (CTAC) PET images. We developed a 3D pix2pix deep learning AC (DLAC) framework via a U-net + ResNet-based generator and a convolutional neural network-based discriminator. Paired static or dynamic NAC and CTAC PET images from 60 rest-only subjects were used as network inputs and labels for static (S-DLAC) and dynamic (D-DLAC) training, respectively. The pre-trained S-DLAC network was then fine-tuned by paired dynamic NAC and CTAC PET frames of 60 rest-only subjects to derive an improved D-DLAC-FT for dynamic PET images. The 14 rest-stress subjects were used as an internal testing dataset and separately tested on different network models without training. The proposed methods were evaluated using visual quality and quantitative metrics. The proposed S-DLAC, D-DLAC, and D-DLAC-FT methods were consistent with clinical CTAC in terms of various images and quantitative metrics. The S-DLAC (slope = 0.9423, R2 = 0.947) showed a higher correlation with the reference static CTAC as compared to static NAC (slope = 0.0992, R2 = 0.654). D-DLAC-FT yielded lower myocardial blood flow (MBF) errors in the whole left ventricular myocardium than D-DLAC, but with no significant difference, both for the 60 rest-state subjects (6.63 ± 5.05% vs. 7.00 ± 6.84%, p = 0.7593) and the 14 stress-state subjects (1.97 ± 2.28% vs. 3.21 ± 3.89%, p = 0.8595). The proposed S-DLAC, D-DLAC, and D-DLAC-FT methods achieve comparable performance with clinical CTAC. Transfer learning shows promising potential for dynamic MP PET.

Artifacts from implanted leads in cardiac PET using CT-based attenuation correction

Magnetic resonance imaging (MRI) has the limitation of low imaging speed. Acceleration methods using under-sampled k-space data have been widely exploited to improve data acquisition without reducing the image quality. Sensitivity encoding (SENSE) is the most commonly used method for multi-channel imaging. However, SENSE has the drawback of severe g-factor artifacts when the under-sampling factor is high. This paper applies generative adversarial networks (GAN) to remove g-factor artifacts from SENSE reconstructions. Our method was evaluated on a public knee database containing 20 healthy participants. We compared our method with conventional GAN using zero-filled (ZF) images as input. Structural similarity (SSIM), peak signal to noise ratio (PSNR), and normalized mean square error (NMSE) were calculated for the assessment of image quality. A paired student's t-test was conducted to compare the image quality metrics between the different methods. Statistical significance was considered at P<0.01. The proposed method outperformed SENSE, variational network (VN), and ZF + GAN methods in terms of SSIM (SENSE + GAN: 0.81±0.06, SENSE: 0.40±0.07, VN: 0.79±0.06, ZF + GAN: 0.77±0.06), PSNR (SENSE + GAN: 31.90±1.66, SENSE: 22.70±1.99, VN: 31.35±2.01, ZF + GAN: 29.95±1.59), and NMSE (×10-7) (SENSE + GAN: 0.95±0.34, SENSE: 4.81±1.33, VN: 0.97±0.30, ZF + GAN: 1.60±0.84) with an under-sampling factor of up to 6-fold. This study demonstrated the feasibility of using GAN to improve the performance of SENSE reconstruction. The improvement of reconstruction is more obvious for higher under-sampling rates, which shows great potential for many clinical applications.

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. 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. 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. 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.

Attenuation correction in cardiac PET: To raise awareness for a problem which is as old as PET/CT

CT-based attenuation correction may influence cardiac PET owing to its higher susceptibility to misalignment compared with conventional (68)Ge transmission scans. The aims of this study were to evaluate whether CT attenuation correction leads to changes in tracer distribution compared with conventional cardiac PET and to determine a suitable CT protocol. A total of 27 patients underwent PET/CT and subsequently a PET scan. Twenty patients received a low-dose CT (LDCT group; 120 kV, 26 mA, 8-s scan time), seven patients a slow CT (SCT group; 120 kV, 99 mA, 46-s scan time) and ten patients an ultra-low-dose CT (ULDCT group; 80 kV, 13 mA, 5-s scan time) as the transmission scan in PET/CT. Polar maps were divided into 17 segments and regression analysis was computed in every scan pair (CT attenuation corrected-(68)Ge attenuation corrected). Correlation coefficient (r), the slope (s) and the offset (os) of the regression line were determined. Visual assessment of misalignment between the transmission and emission data was performed. The effective dose of the different transmission scans was calculated. Overall, there was a moderate correlation between the mean values measured in all segments on PET/CT and on PET when using LDCT (r=0.78, p<0.0001), SCT (r=0.79, p<0.0001) and ULDCT (r=0.82, p<0.0001). No differences were observed when comparing the scores assigned in the visual misalignment assessment in the three groups (p=0.12). The differences between the results from the regression analysis observed in the respective groups were not statistically significant (Kruskal-Wallis p=0.11 for r, p=0.67 for s and p=0.27 for os). The effective dose was lowest for the ULDCT. Our study shows that CT-based attenuation correction is feasible for cardiac PET imaging. The results indicate that ultra-low-dose CT is the preferable choice for transmission scanning.

The goal of this study is to leverage an advanced fast imaging technique, wave-controlled aliasing in parallel imaging (Wave-CAIPI), and a generative adversarial network (GAN) for denoising to achieve accelerated high-quality high-signal-to-noise-ratio (SNR) volumetric magnetic resonance imaging (MRI). Three-dimensional (3D) T2 -weighted fluid-attenuated inversion recovery (FLAIR) image data were acquired on 33 multiple sclerosis (MS) patients using a prototype Wave-CAIPI sequence (acceleration factor R=3×2, 2.75min) and a standard T2 -sampling perfection with application-optimized contrasts by using flip angle evolution (SPACE) FLAIR sequence (R=2, 7.25min). A hybrid denoising GAN entitled "HDnGAN" consisting of a 3D generator and a 2D discriminator was proposed to denoise highly accelerated Wave-CAIPI images. HDnGAN benefits from the improved image synthesis performance provided by the 3D generator and increased training samples from a limited number of patients for training the 2D discriminator. HDnGAN was trained and validated on data from 25 MS patients with the standard FLAIR images as the target and evaluated on data from eight MS patients not seen during training. HDnGAN was compared to other denoising methods including adaptive optimized nonlocal means (AONLM), block matching with 4D filtering (BM4D), modified U-Net (MU-Net), and 3D GAN in qualitative and quantitative analysis of output images using the mean squared error (MSE) and Visual Geometry Group (VGG) perceptual loss compared to standard FLAIR images, and a reader assessment by two neuroradiologists regarding sharpness, SNR, lesion conspicuity, and overall quality. Finally, the performance of these denoising methods was compared at higher noise levels using simulated data with added Rician noise. HDnGAN effectively denoised low-SNR Wave-CAIPI images with sharpness and rich textural details, which could be adjusted by controlling the contribution of the adversarial loss to the total loss when training the generator. Quantitatively, HDnGAN (λ=10-3 ) achieved low MSE and the lowest VGG perceptual loss. The reader study showed that HDnGAN (λ=10-3 ) significantly improved the SNR of Wave-CAIPI images (p<0.001), outperformed AONLM (p=0.015), BM4D (p<0.001), MU-Net (p<0.001), and 3D GAN (λ=10-3 ) (p<0.001) regarding image sharpness, and outperformed MU-Net (p<0.001) and 3D GAN (λ=10-3 ) (p=0.001) regarding lesion conspicuity. The overall quality score of HDnGAN (λ=10-3 ) (4.25±0.43) was significantly higher than those from Wave-CAIPI (3.69±0.46, p=0.003), BM4D (3.50±0.71, p=0.001), MU-Net (3.25±0.75, p<0.001), and 3D GAN (λ=10-3 ) (3.50±0.50, p<0.001), with no significant difference compared to standard FLAIR images (4.38±0.48, p=0.333). The advantages of HDnGAN over other methods were more obvious at higher noise levels. HDnGAN provides robust and feasible denoising while preserving rich textural detail in empirical volumetric MRI data. Our study using empirical patient data and systematic evaluation supports the use of HDnGAN in combination with modern fast imaging techniques such as Wave-CAIPI to achieve high-fidelity fast volumetric MRI and represents an important step to the clinical translation of GANs.

Introduction: Inflammatory cardiomyopathies such as sarcoidosis are associated with increased focal 18F-FDG uptake on cardiac positron emission tomography-computed tomography (PET-CT). It has been reported that patients with genetic cardiomyopathies with desmosomal mutations may also have abnormal cardiac PET scans. The incidence and type of genetic cardiomyopathies with abnormal cardiac PET imaging remains unclear. Objective: To determine the prevalence of genetic cardiomyopathies in patients with abnormal cardiac PET imaging. Methods: Data from patients who underwent cardiac PET imaging at our institution between 2018-2024 were analyzed. Those who completed genetic testing were further identified and evaluated for mutation variant type and imaging findings. Results: Of 605 patients with non-ischemic cardiomyopathy referred for cardiac PET scans, 67 (11%, mean age 51.8 ± 15.4 years, 69% male, mean LVEF 46.5 ± 15.1%) had undergone genetic testing. Of these, 21% (n = 14) had evidence of focal inflammation and 15% (n = 10) had indeterminate PET results with assessment of focal inflammatory findings limited by incomplete suppression of physiologic metabolism at the time of imaging. Of patients with focal inflammation, 36% (n = 5) tested positive for known pathogenic gene variants: DSP (n =1), FLNC (n = 1), LMNA (n = 1), and MYH7 (n = 2). An additional 4 patients with evidence of focal inflammation had variants of uncertain significance (VUS). Among the 10 patients with indeterminate cardiac PET findings, 3 had pathogenic mutations: DSP (n = 1), HFE (n = 1), and TTN (n = 1). There was no difference in the cardiac PET imaging findings of patients with versus without genetic cardiomyopathies with regard to focality or location of FDG uptake. Conclusion: Pathogenic mutations beyond desmosomal gene variants can present with evidence of inflammation on cardiac PET. Genetic testing amongst non-ischemic cardiomyopathy patients with abnormal cardiac PET scans should be considered to evaluate for other etiologies beyond sarcoidosis.

The aims of this study were to assess the performance of FDG PET at PET/CT imaging for the detection of pulmonary metastases and to evaluate differences in lesion detectability on attenuation-corrected (AC) and non-attenuation corrected (NAC) PET images. The institutional PET/CT database was searched for patients with pulmonary metastases of 3-60 mm in diameter. Ninety-two patients with 438 metastases to the lungs were included in the study. The primary tumours were 33 malignant melanomas, 12 carcinomas of unknown primary, 11 colorectal carcinomas, eight differentiated thyroid carcinomas, seven aggressive non-Hodgkin's lymphomas, six head and neck cancers, three breast cancers, two prostate cancers and ten others. Lesion detectability was visually compared between PET and CT and between AC and NAC PET images using a five-point scale. Of the 438 pulmonary metastases, 174 were detected with FDG PET (39.7%), six of them on NAC images only (not significant). Visual scores were higher on NAC images in 41.4% and equal in 54.6% of lesions. The sensitivity of FDG PET increased significantly from 0.405 for metastases of 5-7 mm in diameter to 0.784 for lesions of 8-10 mm and to 0.935 for lesions measuring 11-29 mm in diameter. No metastases smaller than 5 mm in diameter were seen on PET images. FDG PET/CT is useful for the assessment of pulmonary metastases. The frequency of lesion detection is similar for AC and NAC PET images. A reduced sensitivity of FDG PET has to be considered for lesions smaller than 11 mm in diameter.