Non-rigid motion-compensated 3D whole-heart T2 mapping in a hybrid 3T PET-MR system.

PURPOSE: Simultaneous PET-MRI improves inflammatory cardiac disease diagnosis. However, challenges persist in respiratory motion and mis-registration between free-breathing 3D PET and 2D breath-held MR images. We propose a free-breathing non-rigid motion-compensated 3D T2 -mapping sequence enabling whole-heart myocardial tissue characterization in a hybrid 3T PET-MR system and provides non-rigid respiratory motion fields to correct also simultaneously acquired PET data. METHODS: Free-breathing 3D whole-heart T2 -mapping was implemented on a hybrid 3T PET-MRI system. Three datasets were acquired with different T2 -preparation modules (0, 28, 55 ms) using 3-fold undersampled variable-density Cartesian trajectory. Respiratory motion was estimated via virtual 3D image navigators, enabling multi-contrast non-rigid motion-corrected MR reconstruction. T2 -maps were computed using dictionary-matching. Approach was tested in phantom, 8 healthy subjects, 14 MR only and 2 PET-MR patients with suspected cardiac disease and compared with spin echo reference (phantom) and clinical 2D T2 -mapping (in-vivo). RESULTS: Phantom results show a high correlation (R2 = 0.996) between proposed approach and gold standard 2D T2 mapping. In-vivo 3D T2 -mapping average values in healthy subjects (39.0 ± 1.4 ms) and patients (healthy tissue) (39.1 ± 1.4 ms) agree with conventional 2D T2 -mapping (healthy = 38.6 ± 1.2 ms, patients = 40.3 ± 1.7 ms). Bland-Altman analysis reveals bias of 1.8 ms and 95% limits of agreement (LOA) of -2.4-6 ms for healthy subjects, and bias of 1.3 ms and 95% LOA of -1.9 to 4.6 ms for patients. CONCLUSION: Validated efficient 3D whole-heart T2 -mapping at hybrid 3T PET-MRI provides myocardial inflammation characterization and non-rigid respiratory motion fields for simultaneous PET data correction. Comparable T2 values were achieved with both 3D and 2D methods. Improved image quality was observed in the PET images after MR-based motion correction.

Motion-corrected and high-resolution anatomically assisted (MOCHA) reconstruction of arterial spin labeling MRI.

PURPOSE: A model-based reconstruction framework is proposed for motion-corrected and high-resolution anatomically assisted (MOCHA) reconstruction of arterial spin labeling (ASL) data. In this framework, all low-resolution ASL control-label pairs are used to reconstruct a single high-resolution cerebral blood flow (CBF) map, corrected for rigid-motion, point-spread-function blurring and partial volume effect. METHODS: Six volunteers were recruited for CBF imaging using pseudo-continuous ASL labeling, two-shot 3D gradient and spin-echo sequences and high-resolution T1 -weighted MRI. For 2 volunteers, high-resolution scans with double and triple resolution in the partition direction were additionally collected. Simulations were designed for evaluations against a high-resolution ground-truth CBF map, including a simulated hyperperfused lesion and hyperperfusion/hypoperfusion abnormalities. The MOCHA technique was compared with standard reconstruction and a 3D linear regression partial-volume effect correction method and was further evaluated for acquisitions with reduced control-label pairs and k-space undersampling. RESULTS: The MOCHA reconstructions of low-resolution ASL data showed enhanced image quality, particularly in the partition direction. In simulations, both MOCHA and 3D linear regression provided more accurate CBF maps than the standard reconstruction; however, MOCHA resulted in the lowest errors and well delineated the abnormalities. The MOCHA reconstruction of standard-resolution in vivo data showed good agreement with higher-resolution scans requiring 4-times and 9-times longer acquisitions. The MOCHA reconstruction was found to be robust for 4-times-accelerated ASL acquisitions, achieved by reduced control-label pairs or k-space undersampling. CONCLUSION: The MOCHA reconstruction reduces partial-volume effect by direct reconstruction of CBF maps in the high-resolution space of the corresponding anatomical image, incorporating motion correction and point spread function modeling. Following further evaluation, MOCHA should promote the clinical application of ASL.

Multi-modal synergistic PET and MR reconstruction using mutually weighted quadratic priors.

PURPOSE: To propose a framework for synergistic reconstruction of PET-MR and multi-contrast MR data to improve the image quality obtained from noisy PET data and from undersampled MR data. THEORY AND METHODS: Weighted quadratic priors were devised to preserve common boundaries between PET-MR images while reducing noise, PET Gibbs ringing, and MR undersampling artifacts. These priors are iteratively reweighted using normalized multi-modal Gaussian similarity kernels. Synergistic PET-MR reconstructions were built on the PET maximum a posteriori expectation maximization algorithm and the MR regularized sensitivity encoding method. The proposed approach was compared to conventional methods, total variation, and prior-image weighted quadratic regularization methods. Comparisons were performed on a simulated [18 F]fluorodeoxyglucose-PET and T1 /T2 -weighted MR brain phantom, 2 in vivo T1 /T2 -weighted MR brain datasets, and an in vivo [18 F]fluorodeoxyglucose-PET and fluid-attenuated inversion recovery/T1 -weighted MR brain dataset. RESULTS: Simulations showed that synergistic reconstructions achieve the lowest quantification errors for all image modalities compared to conventional, total variation, and weighted quadratic methods. Whereas total variation regularization preserved modality-unique features, this method failed to recover PET details and was not able to reduce MR artifacts compared to our proposed method. For in vivo MR data, our method maintained similar image quality for 3× and 14× accelerated data. Reconstruction of the PET-MR dataset also demonstrated improved performance of our method compared to the conventional independent methods in terms of reduced Gibbs and undersampling artifacts. CONCLUSION: The proposed methodology offers a robust multi-modal synergistic image reconstruction framework that can be readily built on existing established algorithms.

Motion-corrected simultaneous cardiac positron emission tomography and coronary MR angiography with high acquisition efficiency.

PURPOSE: Develop a framework for efficient free-breathing simultaneous whole-heart coronary magnetic resonance angiography (CMRA) and cardiac positron emission tomography (PET) on a 3 Tesla PET-MR system. METHODS: An acquisition that enables nonrigid motion correction of both CMRA and PET has been developed. The proposed method estimates translational motion from low-resolution 2D MR image navigators acquired at each heartbeat and 3D nonrigid respiratory motion between different respiratory bins from the CMRA data itself. Estimated motion is used for correcting the CMRA as well as the emission and attenuation PET data sets to the same respiratory position. The CMRA approach was studied in 10 healthy subjects and compared for both left and right coronary arteries (LCA, RCA) against a reference scan with diaphragmatic navigator gating and tracking. The PET-CMRA approach was tested in 5 oncology patients with 18 F-FDG myocardial uptake. PET images were compared against uncorrected and gated PET reconstructions. RESULTS: For the healthy subjects, no statistically significant differences in vessel length and sharpness (P > 0.01) were observed between the proposed approach and the reference acquisition with navigator gating and tracking, although data acquisition was significantly shorter. The proposed approach improved CMRA vessel sharpness by 37.9% and 49.1% (LCA, RCA) and vessel length by 48.0% and 36.7% (LCA, RCA) in comparison with no motion correction for all the subjects. Motion-corrected PET images showed improved sharpness of the myocardium compared to uncorrected reconstructions and reduced noise compared to gated reconstructions. CONCLUSION: Feasibility of a new respiratory motion-compensated simultaneous cardiac PET-CMRA acquisition has been demonstrated. Magn Reson Med 79:339-350, 2018. © 2017 The Authors Magnetic Resonance in Medicine published by Wiley Periodicals, Inc. on behalf of International Society for Magnetic Resonance in Medicine. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

MR-guided joint reconstruction of activity and attenuation in brain PET-MR.

With the advent of time-of-flight (TOF) PET scanners, joint maximum-likelihood reconstruction of activity and attenuation (MLAA) maps has recently regained attention for the estimation of PET attenuation maps from emission data. However, the estimated attenuation and activity maps are scaled by unknown scaling factors. We recently demonstrated that in hybrid PET-MR, the scaling issue of this algorithm can be effectively addressed by imposing MR spatial constraints on the estimation of attenuation maps using a penalized MLAA (P-MLAA+) algorithm. With the advent of simultaneous PET-MR systems, MRI-guided PET image reconstruction has also gained attention for improving the quantitative accuracy of PET images, usually degraded by noise and partial volume effects. The aim of this study is therefore to increase the benefits of MRI information for improving the quantitative accuracy of PET images by exploiting MRI-based anatomical penalty functions to guide the reconstruction of both activity and attenuation maps during their joint estimation. We employed an anato-functional joint entropy penalty function for the reconstruction of activity and an anatomical quadratic penalty function for the reconstruction of attenuation. The resulting algorithm was referred to as P-MLAA++ since it exploits both activity and attenuation penalty functions. The performance of the P-MLAA algorithms were compared with MLAA and the widely used activity reconstruction algorithms such as maximum likelihood expectation maximization (MLEM) and penalized MLEM (P-MLEM) both corrected for attenuation using a conventional MRI segmentation-based attenuation correction (MRAC) method. The studied methods were evaluated using simulations and clinical studies taking the PET image reconstructed using reference CT-based attenuation maps as a reference. The simulation results showed that the proposed method can notably improve the visual quality of the PET images by reducing noise while preserving structural boundaries and at the same time improving the quantitative accuracy of the PET images. Our clinical reconstruction results showed that the MLEM-MRAC, P-MLEM-MRAC, MLAA, P-MLAA+ and P-MLAA++ algorithms result in, on average, quantification errors of -13.5 ± 3.1%, -13.4 ± 3.1%, -2.0 ± 6.5%, -3.0 ± 3.5% and -4.2 ± 3.6%, respectively, in different regions of the brain. In conclusion, whilst the P-MLAA+ algorithm showed the best overall quantification performance, the proposed P-MLAA++ algorithm provided simultaneous partial volume and attenuation corrections with only a minor compromise of PET quantification.

Metabotropic Glutamate Receptor Type 5 (mGluR5) Cortical Abnormalities in Focal Cortical Dysplasia Identified In Vivo With [11C]ABP688 Positron-Emission Tomography (PET) Imaging.

Metabotropic glutamate receptor type 5 (mGluR5) abnormalities have been described in tissue resected from epilepsy patients with focal cortical dysplasia (FCD). To determine if these abnormalities could be identified in vivo, we investigated mGluR5 availability in 10 patients with focal epilepsy and an MRI diagnosis of FCD using positron-emission tomography (PET) and the radioligand [11C]ABP688. Partial volume corrected [11C]ABP688 binding potentials (BPND) were computed using the cerebellum as a reference region. Each patient was compared to homotopic cortical regions in 33 healthy controls using region-of-interest (ROI) and vertex-wise analyses. Reduced [11C]ABP688 BPND in the FCD was seen in 7/10 patients with combined ROI and vertex-wise analyses. Reduced FCD BPND was found in 4/5 operated patients (mean follow-up: 63 months; Engel I), of whom surgical specimens revealed FCD type IIb or IIa, with most balloon cells showing negative or weak mGluR5 immunoreactivity as compared to their respective neuropil and normal neurons at the border of resections. [11C]ABP688 PET shows for the first time in vivo evidence of reduced mGluR5 availability in FCD, indicating focal glutamatergic alterations in malformations of cortical development, which cannot be otherwise clearly demonstrated through resected tissue analyses.

Characterization of age/sex and the regional distribution of mGluR5 availability in the healthy human brain measured by high-resolution [(11)C]ABP688 PET.

PURPOSE: Metabotropic glutamate receptor type 5 (mGluR5) is a G protein-coupled receptor that has been implicated in several psychiatric and neurological diseases. The radiopharmaceutical [(11)C]ABP688 allows for in vivo quantification of mGluR5 availability using positron emission tomography (PET). In this study, we aimed to detail the regional distribution of [(11)C]ABP688 binding potential (BPND) and the existence of age/sex effects in healthy individuals. METHODS: Thirty-one healthy individuals aged 20 to 77 years (men, n = 18, 45.3 ± 18.2 years; females, n = 13, 41.5 ± 19.6 years) underwent imaging with [(11)C]ABP688 using the high-resolution research tomograph (HRRT). We developed an advanced partial volume correction (PVC) method using surface-based analysis in order to accurately estimate the regional variation of radioactivity. BPND was calculated using the simplified reference tissue model, with the cerebellum as the reference region. Surface-based and volume-based analyses were performed for 39 cortical and subcortical regions of interest per hemisphere. RESULTS: We found the highest [(11)C]ABP688 BPND in the lateral prefrontal and anterior cingulate cortices. The lowest [(11)C]ABP688 BPND was observed in the pre- and post-central gyri as well as the occipital lobes and the thalami. No sex effect was observed. Associations between age and [(11)C]ABP688 BPND without PVC were observed in the right amygdala and left putamen, but were not significant after multiple comparisons correction. CONCLUSIONS: The present results highlight complexities underlying brain adaptations during the aging process, and support the notion that certain aspects of neurotransmission remain stable during the adult life span.

Cocaine cue-induced dopamine release in the human prefrontal cortex.

BACKGROUND: Accumulating evidence indicates that drug-related cues can induce dopamine (DA) release in the striatum of substance abusers. Whether these same cues provoke DA release in the human prefrontal cortex remains unknown. METHODS: We used high-resolution positron emission tomography with [18F]fallypride to measure cortical and striatal DA D2/3 receptor availability in the presence versus absence of drug-related cues in volunteers with current cocaine dependence. RESULTS: Twelve individuals participated in our study. Among participants reporting a craving response (9 of 12), exposure to the cocaine cues significantly decreased [18F]fallypride binding potential (BPND) values in the medial orbitofrontal cortex and striatum. In all 12 participants, individual differences in the magnitude of craving correlated with BPND changes in the medial orbitofrontal cortex, dorsolateral prefrontal cortex, anterior cingulate, and striatum. Consistent with the presence of autoreceptors on mesostriatal but not mesocortical DA cell bodies, midbrain BPND values were significantly correlated with changes in BPND within the striatum but not the cortex. The lower the midbrain D2 receptor levels, the greater the striatal change in BPND and self-reported craving. LIMITATIONS: Limitations of this study include its modest sample size, with only 2 female participants. Newer tracers might have greater sensitivity to cortical DA release. CONCLUSION: In people with cocaine use disorders, the presentation of drug-related cues induces DA release within cortical and striatal regions. Both effects are associated with craving, but only the latter is regulated by midbrain autoreceptors. Together, the results suggest that cortical and subcortical DA responses might both influence drug-focused incentive motivational states, but with separate regulatory mechanisms.

External awareness and GABA--a multimodal imaging study combining fMRI and [18F]flumazenil-PET.

Awareness is an essential feature of the human mind that can be directed internally, that is, toward our self, or externally, that is, toward the environment. The combination of internal and external information is crucial to constitute our sense of self. Although the underlying neuronal networks, the so-called intrinsic and extrinsic systems, have been well-defined, the associated biochemical mechanisms still remain unclear. We used a well-established functional magnetic resonance imaging (fMRI) paradigm for internal (heartbeat counting) and external (tone counting) awareness and combined this technique with [(18)F]FMZ-PET imaging in the same healthy subjects. Focusing on cortical midline regions, the results showed that both stimuli types induce negative BOLD responses in the mPFC and the precuneus. Carefully controlling for structured noise in fMRI data, these results were also confirmed in an independent data sample using the same paradigm. Moreover, the degree of the GABAA receptor binding potential within these regions was correlated with the neuronal activity changes associated with external, rather than internal awareness when compared to fixation. These data support evidence that the inhibitory neurotransmitter GABA is an influencing factor in the differential processing of internally and externally guided awareness. This in turn has implications for our understanding of the biochemical mechanisms underlying awareness in general and its potential impact on psychiatric disorders.

Kinetic model-informed deep learning for multiplexed PET image separation.

BACKGROUND: Multiplexed positron emission tomography (mPET) imaging can measure physiological and pathological information from different tracers simultaneously in a single scan. Separation of the multiplexed PET signals within a single PET scan is challenging due to the fact that each tracer gives rise to indistinguishable 511 keV photon pairs, and thus no unique energy information for differentiating the source of each photon pair. METHODS: Recently, many applications of deep learning for mPET image separation have been concentrated on pure data-driven methods, e.g., training a neural network to separate mPET images into single-tracer dynamic/static images. These methods use over-parameterized networks with only a very weak inductive prior. In this work, we improve the inductive prior of the deep network by incorporating a general kinetic model based on spectral analysis. The model is incorporated, along with deep networks, into an unrolled image-space version of an iterative fully 4D PET reconstruction algorithm. RESULTS: The performance of the proposed method was evaluated on a simulated brain image dataset for dual-tracer [ 18 F]FDG+[ 11 C]MET PET image separation. The results demonstrate that the proposed method can achieve separation performance comparable to that obtained with single-tracer imaging. In addition, the proposed method outperformed the model-based separation methods (the conventional voxel-wise multi-tracer compartment modeling method (v-MTCM) and the image-space dual-tracer version of the fully 4D PET image reconstruction algorithm (IS-F4D)), as well as a pure data-driven separation [using a convolutional encoder-decoder (CED)], with fewer training examples. CONCLUSIONS: This work proposes a kinetic model-informed unrolled deep learning method for mPET image separation. In simulation studies, the method proved able to outperform both the conventional v-MTCM method and a pure data-driven CED with less training data.

Deep learning-based PET image denoising and reconstruction: a review.

This review focuses on positron emission tomography (PET) imaging algorithms and traces the evolution of PET image reconstruction methods. First, we provide an overview of conventional PET image reconstruction methods from filtered backprojection through to recent iterative PET image reconstruction algorithms, and then review deep learning methods for PET data up to the latest innovations within three main categories. The first category involves post-processing methods for PET image denoising. The second category comprises direct image reconstruction methods that learn mappings from sinograms to the reconstructed images in an end-to-end manner. The third category comprises iterative reconstruction methods that combine conventional iterative image reconstruction with neural-network enhancement. We discuss future perspectives on PET imaging and deep learning technology.

AI for PET image reconstruction.

Image reconstruction for positron emission tomography (PET) has been developed over many decades, with advances coming from improved modelling of the data statistics and improved modelling of the imaging physics. However, high noise and limited spatial resolution have remained issues in PET imaging, and state-of-the-art PET reconstruction has started to exploit other medical imaging modalities (such as MRI) to assist in noise reduction and enhancement of PET's spatial resolution. Nonetheless, there is an ongoing drive towards not only improving image quality, but also reducing the injected radiation dose and reducing scanning times. While the arrival of new PET scanners (such as total body PET) is helping, there is always a need to improve reconstructed image quality due to the time and count limited imaging conditions. Artificial intelligence (AI) methods are now at the frontier of research for PET image reconstruction. While AI can learn the imaging physics as well as the noise in the data (when given sufficient examples), one of the most common uses of AI arises from exploiting databases of high-quality reference examples, to provide advanced noise compensation and resolution recovery. There are three main AI reconstruction approaches: (i) direct data-driven AI methods which rely on supervised learning from reference data, (ii) iterative (unrolled) methods which combine our physics and statistical models with AI learning from data, and (iii) methods which exploit AI with our known models, but crucially can offer benefits even in the absence of any example training data whatsoever. This article reviews these methods, considering opportunities and challenges of AI for PET reconstruction.

Synthesized Image Reconstruction for Post-Reconstruction Resolution Recovery.

Resolution recovery (RR) techniques in positron emission tomography (PET) imaging aim to mitigate spatial resolution losses and related inaccuracies in quantification by using a model of the system's point spread function (PSF) during reconstruction or post-processing. However, including PSF modeling in fully 3-D image reconstruction is far from trivial as access to the scanner-specific forward and back-projectors is required, along with access to the 3-D sinogram data. Hence, post-reconstruction RR methods, such as the Richardson-Lucy (RL) algorithm, can be more practical. However, the RL method leads to relatively rapid noise amplification in early image iterations, giving inferior image quality compared to iterates obtained by placing the PSF model in the reconstruction algorithm. We propose a post-reconstruction RR method by synthesizing PET data by a forward projection of an initial real data reconstruction (such reconstructions are usually available via a scanner's standard reconstruction software). The synthetic PET data are then used to reconstruct an image, but crucially now including a modeled PSF within the system model used during reconstruction. Results from simulations and real data demonstrate the proposed method improves image quality compared to the RL algorithm, whilst avoiding the need for scanner-specific projectors and raw sinogram data (as required by standard PSF modeling within reconstruction).

Deep Learned Segmentations of Inflammation for Novel 99mTc-maraciclatide Imaging of Rheumatoid Arthritis.

Rheumatoid arthritis (RA) is an autoimmune disease that causes joint pain, stiffness, and erosion. Power Doppler ultrasound and MRI are imaging modalities used in detecting and monitoring the disease, but they have limitations. 99mTc-maraciclatide gamma camera imaging is a novel technique that can detect joint inflammation at all sites in a single examination and has been shown to correlate with power Doppler ultrasound. In this work, we investigate if machine learning models can be used to automatically segment regions of normal, low, and highly inflamed tissue from 192 99mTc-maraciclatide scans of the hands and wrists from 48 patients. Two models were trained: a thresholding model that learns lower and upper threshold values and a neural-network-based nnU-Net model that uses a convolutional neural network (CNN). The nnU-Net model showed 0.94 ± 0.01, 0.51 ± 0.14, and 0.76 ± 0.16 modified Dice scores for segmenting the normal, low, and highly inflamed tissue, respectively, when compared to clinical segmented labels. This outperforms the thresholding model, which achieved modified Dice scores of 0.92 ± 0.01, 0.14 ± 0.07, and 0.35 ± 0.21, respectively. This is an important first step in developing artificial intelligence (AI) tools to assist clinicians' workflow in the use of this new radiopharmaceutical.

Pre-training via Transfer Learning and Pretext Learning a Convolutional Neural Network for Automated Assessments of Clinical PET Image Quality.

Positron emission tomography (PET) using a fraction of the usual injected dose would reduce the amount of radioligand needed, as well as the radiation dose to patients and staff, but would compromise reconstructed image quality. For performing the same clinical tasks with such images, a clinical (rather than numerical) image quality assessment is essential. This process can be automated with convolutional neural networks (CNNs). However, the scarcity of clinical quality readings is a challenge. We hypothesise that exploiting easily available quantitative information in pretext learning tasks or using established pre-trained networks could improve CNN performance for predicting clinical assessments with limited data. CNNs were pre-trained to predict injected dose from image patches extracted from eight real patient datasets, reconstructed using between 0.5%-100% of the available data. Transfer learning with seven different patients was used to predict three clinically-scored quality metrics ranging from 0-3: global quality rating, pattern recognition and diagnostic confidence. This was compared to pre-training via a VGG16 network at varying pre-training levels. Pre-training improved test performance for this task: the mean absolute error of 0.53 (compared to 0.87 without pre-training), was within clinical scoring uncertainty. Future work may include using the CNN for novel reconstruction methods performance assessment.

Deep Learning Signal Discrimination for Improved Sensitivity at High Specificity for CMOS Intraoperative Probes.

The challenge in delineating the boundary between cancerous and healthy tissue during cancer resection surgeries can be addressed with the use of intraoperative probes to detect cancer cells labeled with radiotracers to facilitate excision. In this study, deep learning algorithms for background gamma ray signal rejection were explored for an intraoperative probe utilizing CMOS monolithic active pixel sensors optimized toward the detection of internal conversion electrons from [Formula: see text]Tc. Two methods utilizing convolutional neural networks (CNNs) were explored for beta-gamma discrimination: 1) classification of event clusters isolated from the sensor array outputs (SAOs) from the probe and 2) semantic segmentation of event clusters within an acquisition frame of an SAO which provides spatial information on the classification. The feasibility of the methods in this study was explored with several radionuclides including 14C, 57Co, and [Formula: see text]Tc. Overall, the classification deep network is able to achieve an improved area under the curve (AUC) of the receiver operating characteristic (ROC), giving 0.93 for 14C beta and [Formula: see text]Tc gamma clusters, compared to 0.88 for a more conventional feature-based discriminator. Further optimization of the lower left region of the ROC by using a customized AUC loss function during training led to an improvement of 31% in sensitivity at low false positive rates compared to the conventional method. The segmentation deep network is able to achieve a mean dice score of 0.93. Through the direct comparison of all methods, the classification method was found to have a better performance in terms of the AUC.

Memory-Efficient Training for Fully Unrolled Deep Learned PET Image Reconstruction with Iteration-Dependent Targets.

We propose a new version of the forward-backward splitting expectation-maximisation network (FBSEM-Net) along with a new memory-efficient training method enabling the training of fully unrolled implementations of 3D FBSEM-Net. FBSEM-Net unfolds the maximum a posteriori expectation-maximisation algorithm and replaces the regularisation step by a residual convolutional neural network. Both the gradient of the prior and the regularisation strength are learned from training data. In this new implementation, three modifications of the original framework are included. First, iteration-dependent networks are used to have a customised regularisation at each iteration. Second, iteration-dependent targets and losses are introduced so that the regularised reconstruction matches the reconstruction of noise-free data at every iteration. Third, sequential training is performed, making training of large unrolled networks far more memory efficient and feasible. Since sequential training permits unrolling a high number of iterations, there is no need for artificial use of the regularisation step as a leapfrogging acceleration. The results obtained on 2D and 3D simulated data show that FBSEM-Net using iteration-dependent targets and losses improves the consistency in the optimisation of the network parameters over different training runs. We also found that using iteration-dependent targets increases the generalisation capabilities of the network. Furthermore, unrolled networks using iteration-dependent regularisation allowed a slight reduction in reconstruction error compared to using a fixed regularisation network at each iteration. Finally, we demonstrate that sequential training successfully addresses potentially serious memory issues during the training of deep unrolled networks. In particular, it enables the training of 3D fully unrolled FBSEM-Net, not previously feasible, by reducing the memory usage by up to 98% compared to a conventional end-to-end training. We also note that the truncation of the backpropagation (due to sequential training) does not notably impact the network's performance compared to conventional training with a full backpropagation through the entire network.

Micro-Networks for Robust MR-Guided Low Count PET Imaging.

Noise suppression is particularly important in low count positron emission tomography (PET) imaging. Post-smoothing (PS) and regularization methods which aim to reduce noise also tend to reduce resolution and introduce bias. Alternatively, anatomical information from another modality such as magnetic resonance (MR) imaging can be used to improve image quality. Convolutional neural networks (CNNs) are particularly well suited to such joint image processing, but usually require large amounts of training data and have mostly been applied outside the field of medical imaging or focus on classification and segmentation, leaving PET image quality improvement relatively understudied. This article proposes the use of a relatively low-complexity CNN (micro-net) as a post-reconstruction MR-guided image processing step to reduce noise and reconstruction artefacts while also improving resolution in low count PET scans. The CNN is designed to be fully 3-D, robust to very limited amounts of training data, and to accept multiple inputs (including competitive denoising methods). Application of the proposed CNN on simulated low (30 M) count data (trained to produce standard (300 M) count reconstructions) results in a 36% lower normalized root mean squared error (NRMSE, calculated over ten realizations against the ground truth) compared to maximum-likelihood expectation maximization (MLEM) used in clinical practice. In contrast, a decrease of only 25% in NRMSE is obtained when an optimized (using knowledge of the ground truth) PS is performed. A 26% NRMSE decrease is obtained with both RM and optimized PS. Similar improvement is also observed for low count real patient datasets. Overfitting to training data is demonstrated to occur as the network size is increased. In an extreme case, a U-net (which produces better predictions for training data) is shown to completely fail on test data due to overfitting to this case of very limited training data. Meanwhile, the resultant images from the proposed CNN (which has low training data requirements) have lower noise, reduced ringing, and partial volume effects, as well as sharper edges and improved resolution compared to conventional MLEM.

MR-guided motion-corrected PET image reconstruction for cardiac PET-MR.

Simultaneous PET-MR imaging has shown potential for the comprehensive assessment of myocardial health from a single examination. Furthermore, MR-derived respiratory motion information has been shown to improve PET image quality by incorporating this information into the PET image reconstruction. Separately, MR-based anatomically guided PET image reconstruction has been shown to perform effective denoising, but this has been so far demonstrated mainly in brain imaging. To date the combined benefits of motion compensation and anatomical guidance have not been demonstrated for myocardial PET-MR imaging. This work addresses this by proposing a single cardiac PET-MR image reconstruction framework which fully utilises MR-derived information to allow both motion compensation and anatomical guidance within the reconstruction. Methods: Fifteen patients underwent a 18F-FDG cardiac PET-MR scan with a previously introduced acquisition framework. The MR data processing and image reconstruction pipeline produces respiratory motion fields and a high-resolution respiratory motion-corrected MR image with good tissue contrast. This MR-derived information was then included in a respiratory motion-corrected, cardiac-gated, anatomically guided image reconstruction of the simultaneously acquired PET data. Reconstructions were evaluated by measuring myocardial contrast and noise and compared to images from several comparative intermediate methods using the components of the proposed framework separately. Results: Including respiratory motion correction, cardiac gating, and anatomical guidance significantly increased contrast. In particular, myocardium-to-blood pool contrast increased by 143% on average (p<0.0001) compared to conventional uncorrected, non-guided PET images. Furthermore, anatomical guidance significantly reduced image noise compared to non-guided image reconstruction by 16.1% (p<0.0001). Conclusion: The proposed framework for MR-derived motion compensation and anatomical guidance of cardiac PET data was shown to significantly improve image quality compared to alternative reconstruction methods. Each component of the reconstruction pipeline was shown to have a positive impact on the final image quality. These improvements have the potential to improve clinical interpretability and diagnosis based on cardiac PET-MR images.

A PET supersets data framework for exploitation of known motion in image reconstruction.

PURPOSE: Motion during PET data acquisition is either introduced by design (e.g., couch wobble to enrich data sampling) or through unintentional motion of the object under study (e.g., subject head motion). Prior to this present work, such effects have been considered within distinctly different frameworks in PET imaging, with a rather basic approach having been devised for the cases of intentional motion (such as couch wobble or bed translation) in contrast to the relatively advanced modeling approaches devised for object motion. This article unifies the treatment of these different types of acquisition motion within a generalized framework through the use of the PET supersets data format. From this general framework, a range of conventional motion-compensation methods can be examined as special cases, permitting a revealing investigation into both the suboptimality of commonly adopted approximations, as well as the beneficial impact of known acquisition motion on image quality. METHODS: The PET superset data format is a data representation ideally suited for the (potentially lossless) combination of conventional PET data with complementary motion information. Different practical implementations for the reconstruction from PET superset data involving varying levels of approximation to facilitate computation are identified and their impact on the final reconstructed image quality is assessed. Three main simulated case studies are investigated: (i) Motion compensation for subject head motion for 18F-FDG imaging, (ii) motion exploitation using either couch wobble or random motion, and (iii) motion exploitation using involuntary object motion both with and without couch wobble occurring. The motion exploitation case study goes beyond merely correcting for motion: The motion is directly modeled by the iterative reconstruction to exploit the increased sampling which is available to the benefit of reconstructed image quality. RESULTS: Reconstruction from superset data was successfully demonstrated for the case of motion correction for a 18F-FDG brain phantom simulation. Building on this success, the methodology was then applied to the case of motion exploitation. The case study resulted in three important findings. First, only reconstruction implementations which model the motion directly within the iterative reconstruction can succeed in significantly improving image resolution and contrast recovery for a given reconstructed noise level. Therefore, the commonly adopted approximation of binning the motion-adjusted data (which is the only method reported to date in the literature) is suboptimal and underestimates the beneficial impact of methods such as wobble during acquisition. Second, similar improvements were found for both types of motion patterns: Periodic wobble motion as well as random motion. Finally, for the case of simulated realistic involuntary object motion, similar resolution improvements were found (both with and without couch wobble). CONCLUSIONS: The proposed superset framework allows comprehensive analysis of the commonly adopted approximations when considering motion in PET reconstruction. The findings demonstrate that the supersets data format successfully unify reconstruction in the presence of different sources of known acquisition motion into one framework, and as a result leads to hitherto unreported image quality improvements for all the cases tested where the object motion is accurately known.