Clinical Positron Emission Tomography Research Articles

Design and characterization of a hybrid PET detector with DOI capability.

Monolithic or semi-monolithic detectors are attractive for positron emission tomography (PET) scanners with depth-of-interaction (DOI) capability. However, they often require complicated calibrations to determine the interaction positions of gamma photons. We introduce a novel hybrid detector design that combines pixelated and semi-monolithic elements to achieve DOI capability while simplifying the calibrations for positioning. A prototype detector with eight hybrid lutetium-yttrium oxyorthosilicate (LYSO)layers having dimensions of 25.8×12.9×15mm3 was constructed. The energy-weighted and energy-squared weighted averages were used for estimating the x- (pixelated direction) and y-positions (non-pixelated direction). Pseudo-pixels were defined as discrete areas on the flood image based on the crystal look-up table (LUT). The intrinsic spatial resolutions in the pixelated and non-pixelated directions were measured. The ratio of the maximum to the sum of the multipixel photon counter (MPPC) signals was used to estimate the DOI positions. The coincidence timing resolution (CTR) was measured using the average and energy-weighted average of the earliest n time stamps. Two energy windows of 250-700 and 400-600keV were applied for the measurements. The pattern of the flood images showed discrete event clusters, demonstrating that simple calibrations for determining the x- and y-positions of events could be achieved. Under 400-600keV energy window, the average intrinsic spatial resolutions were 1.15 and 1.34mm for the pixelated and non-pixelated directions; the average DOI resolution of the second row of pseudo-pixels was 5.1mm in full width at half maximum (FWHM); when using the energy-weighted average of the earliest four-time stamps, the best CTR of 350ps was achieved. Applying a broader energy window of 250-700keV only slightly degrades the DOI resolution while maintaining the intrinsic resolution; the best CTR degrades to 410ps. The proposed hybrid detector concept was verified, and a prototype detector showed high performance for 3D positioning and timing resolution. The novel detector concept shows promise for preclinical and clinical PET scanners with DOI capability.

Deep learning based bilateral filtering for edge-preserving denoising of respiratory-gated PET

BackgroundResidual image noise is substantial in positron emission tomography (PET) and one of the factors limiting lesion detection, quantification, and overall image quality. Thus, improving noise reduction remains of considerable interest. This is especially true for respiratory-gated PET investigations. The only broadly used approach for noise reduction in PET imaging has been the application of low-pass filters, usually Gaussians, which however leads to loss of spatial resolution and increased partial volume effects affecting detectability of small lesions and quantitative data evaluation. The bilateral filter (BF) — a locally adaptive image filter — allows to reduce image noise while preserving well defined object edges but manual optimization of the filter parameters for a given PET scan can be tedious and time-consuming, hampering its clinical use. In this work we have investigated to what extent a suitable deep learning based approach can resolve this issue by training a suitable network with the target of reproducing the results of manually adjusted case-specific bilateral filtering.MethodsAltogether, 69 respiratory-gated clinical PET/CT scans with three different tracers ([18F]\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$[^{18}\ ext {F}]$$\\end{document}FDG, [18F]\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$[^{18}\ ext {F}]$$\\end{document}L-DOPA, [68Ga]\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$[^{68}\ ext {Ga}]$$\\end{document}DOTATATE) were used for the present investigation. Prior to data processing, the gated data sets were split, resulting in a total of 552 single-gate image volumes. For each of these image volumes, four 3D ROIs were delineated: one ROI for image noise assessment and three ROIs for focal uptake (e.g. tumor lesions) measurements at different target/background contrast levels. An automated procedure was used to perform a brute force search of the two-dimensional BF parameter space for each data set to identify the “optimal” filter parameters to generate user-approved ground truth input data consisting of pairs of original and optimally BF filtered images. For reproducing the optimal BF filtering, we employed a modified 3D U-Net CNN incorporating residual learning principle. The network training and evaluation was performed using a 5-fold cross-validation scheme. The influence of filtering on lesion SUV quantification and image noise level was assessed by calculating absolute and fractional differences between the CNN, manual BF, or original (STD) data sets in the previously defined ROIs.ResultsThe automated procedure used for filter parameter determination chose adequate filter parameters for the majority of the data sets with only 19 patient data sets requiring manual tuning. Evaluation of the focal uptake ROIs revealed that CNN as well as BF based filtering essentially maintain the focal SUVmax\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ ext {SUV}_\ ext {max}$$\\end{document} values of the unfiltered images with a low mean ± SD difference of δSUVmaxCNN,STD\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\delta \ ext {SUV}_\ ext {max}^{\ ext {CNN},\ ext {STD}}$$\\end{document} = (−3.9 ± 5.2)% and δSUVmaxBF,STD\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\delta \ ext {SUV}_\ ext {max}^{\ ext {BF},\ ext {STD}}$$\\end{document} = (−4.4 ± 5.3)%. Regarding relative performance of CNN versus BF, both methods lead to very similar SUVmax\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ ext {SUV}_\ ext {max}$$\\end{document} values in the vast majority of cases with an overall average difference of δSUVmaxCNN,BF\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\delta \ ext {SUV}_\ ext {max}^{\ ext {CNN},\ ext {BF}}$$\\end{document} = (0.5 ± 4.8)%. Evaluation of the noise properties showed that CNN filtering mostly satisfactorily reproduces the noise level and characteristics of BF with δNoiseCNN,BF\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\delta \ ext {Noise}^{\ ext {CNN},\ ext {BF}}$$\\end{document} = (5.6 ± 10.5)%. No significant tracer dependent differences between CNN and BF were observed.ConclusionsOur results show that a neural network based denoising can reproduce the results of a case by case optimized BF in a fully automated way. Apart from rare cases it led to images of practically identical quality regarding noise level, edge preservation, and signal recovery. We believe such a network might proof especially useful in the context of improved motion correction of respiratory-gated PET studies but could also help to establish BF-equivalent edge-preserving CNN filtering in clinical PET since it obviates time consuming manual BF parameter tuning.

Fully Automated Cassette-Based Synthesis of 2-Deoxy-2-[18F]Fluorocellobiose Using Trasis AllInOne Module.

Due to the continuous rise in global incidence and severity of invasive fungal infections (IFIs), particularly among immunocompromised and immunodeficient patients, there is an urgent demand for swift and accurate fungal pathogen diagnosis. Therefore, the need for fungal-specific positron emission tomography (PET) imaging agents that can detect the infection in the early stages is increasing. Cellobiose, a disaccharide, is readily metabolized by fungal pathogens such as Aspergillus species. Recently, our group reported fluorine-18 labeled cellobiose, 2-deoxy-2-[18F]fluorocellobiose ([18F]FCB), for specific imaging of Aspergillus infection. The positive imaging findings with very low background signal on delayed imaging make this ligand a promising fungal-specific imaging ligand. Inspired by this result, the decision was made to automate the radiolabeling procedure for better reproducibility and to facilitate clinical translation. A Trasis AllInOne (Trasis AIO) automated module was used for this purpose. The reagent vials contain commercially available 2-deoxy-2-[18F]fluoroglucose ([18F]FDG), glucose-1-phosphate, and enzyme (cellobiose phosphorylase). A Sep-Pak cartridge was used to purify the tracer. The overall radiochemical yield was 50%-70% (n = 6, decay corrected) in 75-min synthesis time with a radiochemical purity of > 98%. This is a highly reliable protocol to produce current good manufacturing practice (cGMP)-compliant [18F]FCB for clinical PET imaging.

Fully Automated Cassette-Based Synthesis of 2-Deoxy-2-[18F]Fluorocellobiose Using Trasis AllInOne Module.

Due to the continuous rise in global incidence and severity of invasive fungal infections (IFIs), particularly among immunocompromised and immunodeficient patients, there is an urgent demand for swift and accurate fungal pathogen diagnosis. Therefore, the need for fungal-specific positron emission tomography (PET) imaging agents that can detect the infection in the early stages is increasing. Cellobiose, a disaccharide, is readily metabolized by fungal pathogens such as Aspergillus species. Recently, our group reported fluorine-18 labeled cellobiose, 2-deoxy-2-[18F]fluorocellobiose ([18F]FCB), for specific imaging of Aspergillus infection. The positive imaging findings with very low background signal on delayed imaging make this ligand a promising fungal-specific imaging ligand. Inspired by this result, the decision was made to automate the radiolabeling procedure for better reproducibility and to facilitate clinical translation. A Trasis AllInOne (Trasis AIO) automated module was used for this purpose. The reagent vials contain commercially available 2-deoxy-2-[18F]fluoroglucose ([18F]FDG), glucose-1-phosphate, and enzyme (cellobiose phosphorylase). A Sep-Pak cartridge was used to purify the tracer. The overall radiochemical yield was 50%-70% (n = 6, decay corrected) in 75-min synthesis time with a radiochemical purity of > 98%. This is a highly reliable protocol to produce current good manufacturing practice (cGMP)-compliant [18F]FCB for clinical PET imaging.

Anatomically Guided PET Image Reconstruction Using Conditional Weakly-Supervised Multi-Task Learning Integrating Self-Attention.

To address the lack of high-quality training labels in positron emission tomography (PET) imaging, weakly-supervised reconstruction methods that generate network-based mappings between prior images and noisy targets have been developed. However, the learned model has an intrinsic variance proportional to the average variance of the target image. To suppress noise and improve the accuracy and generalizability of the learned model, we propose a conditional weakly-supervised multi-task learning (MTL) strategy, in which an auxiliary task is introduced serving as an anatomical regularizer for the PET reconstruction main task. In the proposed MTL approach, we devise a novel multi-channel self-attention (MCSA) module that helps learn an optimal combination of shared and task-specific features by capturing both local and global channel-spatial dependencies. The proposed reconstruction method was evaluated on NEMA phantom PET datasets acquired at different positions in a PET/CT scanner and 26 clinical whole-body PET datasets. The phantom results demonstrate that our method outperforms state-of-the-art learning-free and weakly-supervised approaches obtaining the best noise/contrast tradeoff with a significant noise reduction of approximately 50.0% relative to the maximum likelihood (ML) reconstruction. The patient study results demonstrate that our method achieves the largest noise reductions of 67.3% and 35.5% in the liver and lung, respectively, as well as consistently small biases in 8 tumors with various volumes and intensities. In addition, network visualization reveals that adding the auxiliary task introduces more anatomical information into PET reconstruction than adding only the anatomical loss, and the developed MCSA can abstract features and retain PET image details.

DOCK-PET: database of CNS kinetic parameters in the healthy human brain for existing PET tracers.

Information about developed positron emission tomography (PET) tracers and obtained clinical PET images is publicly available in a database. However, findings regarding the kinetic parameters of PET tracers are yet to be summarized. Therefore, in this study, we created an open-access database of central nervous system (CNS) kinetic parameters in the healthy human brain for existing PET tracers (DOCK-PET). Our database includes information on the kinetic parameters and compounds of existing CNS-PET tracers. The kinetic parameter dataset comprises the analysis methods, VT, BPND, K parameters, relevant literature, and study details. The list of PET tracers and kinetic parameter information was compiled through keyword-based searches of PubMed and the Molecular Imaging and Contrast Agent Database (MICAD). The kinetic parameters obtained, including VT, BPND, and K parameters, were reorganized based on the defined brain anatomical regions. All data were rigorously double-checked before being summarized in Microsoft Excel and JavaScript Object Notation (JSON) formats. Of the 247 PET tracers identified through searches using the PubMed and MICAD websites, the kinetic parameters of 120 PET tracers were available. Among the 120 PET tracers, compound structures with chemical and physical properties were obtained from the PubChem website or the ChemDraw software. Furthermore, the affinity information of the 104 PET tracers was gathered from PubChem or extensive literature surveys of the 120 PET tracers. We developed a comprehensive open-access database, DOCK-PET, that includes both kinetic parameters of healthy humans and compound information for existing CNS-PET tracers.

Production of carbon-11 for PET preclinical imaging using a high-repetition rate laser-driven proton source

Most advanced medical imaging techniques, such as positron-emission tomography (PET), require tracers that are produced in conventional particle accelerators. This paper focuses on the evaluation of a potential alternative technology based on laser-driven ion acceleration for the production of radioisotopes for PET imaging. We report for the first time the use of a high-repetition rate, ultra-intense laser system for the production of carbon-11 in multi-shot operation. Proton bunches with energies up to 10–14 MeV were systematically accelerated in long series at pulse rates between 0.1 and 1 Hz using a PW-class laser. These protons were used to activate a boron target via the 11\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{11}$$\\end{document}B(p,n)11\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^{11}$$\\end{document}C nuclear reaction. A peak activity of 234 kBq was obtained in multi-shot operation with laser pulses with an energy of 25 J. Significant carbon-11 production was also achieved for lower pulse energies. The experimental carbon-11 activities measured in this work are comparable to the levels required for preclinical PET, which would be feasible by operating at the repetition rate of current state-of-the-art technology (10 Hz). The scalability of next-generation laser-driven accelerators in terms of this parameter for sustained operation over time could increase these overall levels into the clinical PET range.

Recent developments and challenges in positron emission tomography imaging of gliosis in chronic neuropathic pain.

Understanding the mechanisms that underpin the transition from acute to chronic pain is critical for the development of more effective and targeted treatments. There is growing interest in the contribution of glial cells to this process, with cross-sectional preclinical studies demonstrating specific changes in these cell types capturing targeted timepoints from the acute phase and the chronic phase. In vivo longitudinal assessment of the development and evolution of these changes in experimental animals and humans has presented a significant challenge. Recent technological advances in preclinical and clinical positron emission tomography, including the development of specific radiotracers for gliosis, offer great promise for the field. These advances now permit tracking of glial changes over time and provide the ability to relate these changes to pain-relevant symptomology, comorbid psychiatric conditions, and treatment outcomes at both a group and an individual level. In this article, we summarize evidence for gliosis in the transition from acute to chronic pain and provide an overview of the specific radiotracers available to measure this process, highlighting their potential, particularly when combined with ex vivo/in vitro techniques, to understand the pathophysiology of chronic neuropathic pain. These complementary investigations can be used to bridge the existing gap in the field concerning the contribution of gliosis to neuropathic pain and identify potential targets for interventions.

TAI-GAN: A Temporally and Anatomically Informed Generative Adversarial Network for early-to-late frame conversion in dynamic cardiac PET inter-frame motion correction

Inter-frame motion in dynamic cardiac positron emission tomography (PET) using rubidium-82 (82Rb) myocardial perfusion imaging impacts myocardial blood flow (MBF) quantification and the diagnosis accuracy of coronary artery diseases. However, the high cross-frame distribution variation due to rapid tracer kinetics poses a considerable challenge for inter-frame motion correction, especially for early frames where intensity-based image registration techniques often fail. To address this issue, we propose a novel method called Temporally and Anatomically Informed Generative Adversarial Network (TAI-GAN) that utilizes an all-to-one mapping to convert early frames into those with tracer distribution similar to the last reference frame. The TAI-GAN consists of a feature-wise linear modulation layer that encodes channel-wise parameters generated from temporal information and rough cardiac segmentation masks with local shifts that serve as anatomical information. Our proposed method was evaluated on a clinical 82Rb PET dataset, and the results show that our TAI-GAN can produce converted early frames with high image quality, comparable to the real reference frames. After TAI-GAN conversion, the motion estimation accuracy and subsequent myocardial blood flow (MBF) quantification with both conventional and deep learning-based motion correction methods were improved compared to using the original frames. The code is available at https://github.com/gxq1998/TAI-GAN.

Diagnostic Positron Emission Tomography Imaging with Zirconium-89 Desferrioxamine B Squaramide: From Bench to Bedside.

Molecular imaging with antibodies radiolabeled with positron-emitting radionuclides combines the affinity and selectivity of antibodies with the sensitivity of Positron Emission Tomography (PET). PET imaging allows the visualization and quantification of the biodistribution of the injected radiolabeled antibody, which can be used to characterize specific biological interactions in individual patients. This characterization can provide information about the engagement of the antibody with a molecular target such as receptors present in elevated levels in tumors as well as providing insight into the distribution and clearance of the antibody. Potential applications of clinical PET with radiolabeled antibodies include identifying patients for targeted therapies, characterization of heterogeneous disease, and monitoring treatment response.Antibodies often take several days to clear from the blood pool and localize in tumors, so PET imaging with radiolabeled antibodies requires the use of a radionuclide with a similar radioactive half-life. Zirconium-89 is a positron-emitting radionuclide that has a radioactive half-life of 78 h and relatively low positron emission energy that is well suited to radiolabeling antibodies. It is essential that the zirconium-89 radionuclide be attached to the antibody through chemistry that provides an agent that is stable in vivo with respect to the dissociation of the radionuclide without compromising the biological activity of the antibody.This Account focuses on our research using a simple derivative of the bacterial siderophore desferrioxamine (DFO) with a squaramide ester functional group, DFO-squaramide (DFOSq), to link the chelator to antibodies. In our work, we produce conjugates with an average ∼4 chelators per antibody, and this does not compromise the binding of the antibody to the target. The resulting antibody conjugates of DFOSq are stable and can be easily radiolabeled with zirconium-89 in high radiochemical yields and purity. Automated methods for the radiolabeling of DFOSq-antibody conjugates have been developed to support multicenter clinical trials. Evaluation of several DFOSq conjugates with antibodies and low molecular weight targeting agents in tumor mouse models gave PET images with high tumor uptake and low background. The promising preclinical results supported the translation of this chemistry to human clinical trials using two different radiolabeled antibodies. The potential clinical impact of these ongoing clinical trials is discussed.The use of DFOSq to radiolabel relatively low molecular weight targeting molecules, peptides, and peptide mimetics is also presented. Low molecular weight molecules typically clear the blood pool and accumulate in target tissue more rapidly than antibodies, so they are usually radiolabeled with positron-emitting radionuclides with shorter radioactive half-lives such as fluorine-18 (t1/2 ∼ 110 min) or gallium-68 (t1/2 ∼ 68 min). Radiolabeling peptides and peptide mimetics with zirconium-89, with its longer radioactive half-life (t1/2 = 78 h), could facilitate the centralized manufacture and distribution of radiolabeled tracers. In addition, the ability to image patients at later time points with zirconium-89 based agents (e.g. 4-24 h after injection) may also allow the delineation of small or low-uptake disease sites as the delayed imaging results in increased clearance of the tracer from nontarget tissue and lower background signal.

Modelization of extended axial field-of-view PET scanners to analyze the performance improvement

Positron Emission Tomography (PET) offers non-invasive insights into physiological processes. However, in clinical practice conventional PET scanners face limitations in comprehensive imaging due to their restricted axial field of view (FOV).This study explores the benefits and challenges of extending PET scanner AFOV using Monte Carlo simulation with GATE 9.1. The Biograph Vision Quadra scanner is a novel system based on the Digital Biograph Vision 600 scanner. Both devices use silicon photomultipliers and 3.2 x 3.2 × 20 mm Lutetium Oxyorthosilicate crystals (LSO). However, the 32 detector rings of the Quadra scanner enable a four-fold expansion in the Axial Field Of View (FOV) compared to Vision. As a result, higher sensitivity, and an enhancement of noise equivalent count rate and image quality are achieved. The purpose of this study is to analyze the effect of increasing the number of detector rings and the axial FOV on the scanner performance. To accomplish this, four different geometries are modeled using GATE (GEANT4 Application for Tomographic Emission). Based on the Biograph Vision 600 model, the length and the number of detector rings are gradually extended to reach the length of the Quadra scanner. Thus, four geometries with 8 (25.6 cm), 16 (51.2 cm), 24 (76.8 cm) and 32 (102.4 cm) detector rings are simulated, keeping the crystal configuration and the dimensions in the transaxial plane. The study evaluates sensitivity, Noise Equivalent Count Rate (NECR), and image quality parameters. Results indicate a significant increase in sensitivity and NECR with FOV extension, underscoring enhanced performance. However, artifact occurrence, particularly concentric circular patterns and background homogeneity changes, accentuates with extended AFOV. Normalization algorithms, particularly component-based normalization, mitigate these artifacts. This study contributes to the understanding of the advantages and limitations of extending PET scanner AFOV, offering insights into the effectiveness of component-based normalization for improving PET image quality.

Self-Supervised Pre-Training for Deep Image Prior-Based Robust PET Image Denoising

Deep image prior (DIP) has been successfully applied to positron emission tomography (PET) image restoration, enabling represent implicit prior using only convolutional neural network architecture without training dataset, whereas the general supervised approach requires massive low-and high-quality PET image pairs. To answer the increased need for PET imaging with DIP, it is indispensable to improve the performance of the underlying DIP itself. Here, we propose a self-supervised pre-training model to improve the DIP-based PET image denoising performance. Our proposed pre-training model acquires transferable and generalizable visual representations from only unlabeled PET images by restoring various degraded PET images in a self-supervised approach. We evaluated the proposed method using clinical brain PET data with various radioactive tracers (18F-florbetapir, 11C-Pittsburgh compound-B, 18F-fluoro-2-deoxy-D-glucose, and 15O-CO2) acquired from different PET scanners. The proposed method using the self-supervised pre-training model achieved robust and state-of-the-art denoising performance while retaining spatial details and quantification accuracy compared to other unsupervised methods and pre-training model. These results highlight the potential that the proposed method is particularly effective against rare diseases and probes and helps reduce the scan time or the radiotracer dose without affecting the patients.

Image quality evaluation for a clinical organ-targeted PET camera.

A newly developed clinical organ-targeted Positron Emission Tomography (PET) system (also known as Radialis PET) is tested with a set of standardized and custom tests previously used to evaluate the performance of Positron Emission Mammography (PEM) systems. Imaging characteristics impacting standardized uptake value (SUV) and detectability of small lesions, namely spatial resolution, linearity, uniformity, and recovery coefficients, are evaluated. In-plane spatial resolution was measured as 2.3 mm ± 0.1 mm, spatial accuracy was 0.1 mm, and uniformity measured with flood field and NEMA NU-4 phantom was 11.7% and 8.3% respectively. Selected clinical images are provided as reference to the imaging capabilities under different clinical conditions such as reduced activity of 2-[fluorine-18]-fluoro-2-deoxy-D-glucose (18F-FDG) and time-delayed acquisitions. SUV measurements were performed for selected clinical acquisitions to demonstrate a capability for quantitative image assessment of different types of cancer including for invasive lobular carcinoma with comparatively low metabolic activity. Quantitative imaging performance assessment with phantoms demonstrates improved contrast recovery and spill-over ratio for this PET technology when compared to other commercial organ-dedicated PET systems with similar spatial resolution. Recovery coefficients were measured to be 0.21 for the 1 mm hot rod and up to 0.89 for the 5 mm hot rod of NEMA NU-4 Image Quality phantom. Demonstrated ability to accurately reconstruct activity in tumors as small as 5 mm suggests that the Radialis PET technology may be well suited for emerging clinical applications such as image guided assessment of response to neoadjuvant systemic treatment (NST) in lesions smaller than 2 cm. Also, our results suggest that, while spatial resolution greatly influences the partial volume effect which degrades contrast recovery, optimized count rate performance and image reconstruction workflow may improve recovery coefficients for systems with comparable spatial resolution. We emphasize that recovery coefficient should be considered as a primary performance metric when a PET system is used for accurate lesion size or radiotracer uptake assessments.

Learning Without Real Data Annotations to Detect Hepatic Lesions in PET Images.

Deep neural networks have been recently applied to lesion identification in fluorodeoxyglucose (FDG) positron emission tomography (PET) images, but they typically rely on a large amount of well-annotated data for model training. This is extremely difficult to achieve for neuroendocrine tumors (NETs), because of low incidence of NETs and expensive lesion annotation in PET images. The objective of this study is to design a novel, adaptable deep learning method, which uses no real lesion annotations but instead low-cost, list mode-simulated data, for hepatic lesion detection in real-world clinical NET PET images. We first propose a region-guided generative adversarial network (RG-GAN) for lesion-preserved image-to-image translation. Then, we design a specific data augmentation module for our list-mode simulated data and incorporate this module into the RG-GAN to improve model training. Finally, we combine the RG-GAN, the data augmentation module and a lesion detection neural network into a unified framework for joint-task learning to adaptatively identify lesions in real-world PET data. The proposed method outperforms recent state-of-the-art lesion detection methods in real clinical 68Ga-DOTATATE PET images, and produces very competitive performance with the target model that is trained with real lesion annotations. With RG-GAN modeling and specific data augmentation, we can obtain good lesion detection performance without using any real data annotations. This study introduces an adaptable deep learning method for hepatic lesion identification in NETs, which can significantly reduce human effort for data annotation and improve model generalizability for lesion detection with PET imaging.

Feasibility study of a SiPM-fiber detector for non-invasive measurement of arterial input function for preclinical and clinical positron emission tomography

Pharmacokinetic positron emission tomography (PET) studies rely on the measurement of the arterial input function (AIF), which represents the time-activity curve of the radiotracer concentration in the blood plasma. Traditionally, obtaining the AIF requires invasive procedures, such as arterial catheterization, which can be challenging, time-consuming, and associated with potential risks. Therefore, the development of non-invasive techniques for AIF measurement is highly desirable. This study presents a detector for the non-invasive measurement of the AIF in PET studies. The detector is based on the combination of scintillation fibers and silicon photomultipliers (SiPMs) which leads to a very compact and rugged device. The feasibility of the detector was assessed through Monte Carlo simulations conducted on mouse tail and human wrist anatomies studying relevant parameters such as energy spectrum, detector efficiency and minimum detectable activity (MDA). The simulations involved the use of 18F and 68Ga isotopes, which exhibit significantly different positron ranges. In addition, several prototypes were built in order to study the different components of the detector including the scintillation fiber, the coating of the fiber, the SiPMs, and the operating configuration. Finally, the simulations were compared with experimental measurements conducted using a tube filled with both 18F and 68Ga to validate the obtained results. The MDA achieved for both anatomies (approximately 1000 kBq/mL for mice and 1 kBq/mL for humans) falls below the peak radiotracer concentrations typically found in PET studies, affirming the feasibility of conducting non-invasive AIF measurements with the fiber detector. The sensitivity for measurements with a tube filled with 18F (68Ga) was 1.2 (2.07) cps/(kBq/mL), while for simulations, it was 2.81 (6.23) cps/(kBq/mL). Further studies are needed to validate these results in pharmacokinetic PET studies.

Sensitive Positron Emission Tomography Imaging of PD-L1 Expression in Human Breast and Lung Carcinoma Xenografts Using the Radiometalated Peptide Ga-68-TRAP-WL12.

Noninvasive imaging of the immune checkpoint protein programmed death ligand 1 (PD-L1; synonyms: CD274, B7-H1) holds great promise to improve patient selection and, thus, response rates for immune checkpoint therapy (ICT) with monoclonal antibodies targeting the PD1/PD-L1 axis. The PD-L1 specific peptide WL12 (cyclo(AcY-(NMe)A-N-P-H-L-Hyp-W-S-W(Me)-(NMe)Nle-(NMe)Nle-O-C)-G-NH2) was functionalized with the Gallium-68 chelator TRAP by means of click chemistry (CuAAC). The resulting conjugate TRAP-WL12 was labeled with Gallium-68 within 16 min, with approximately 90% radiochemical yield and 99% radiochemical purity, affording Ga-68-TRAP-WL12 with molar activities typically exceeding 100 MBq/nmol. This radiotracer was characterized by positron emission tomography (PET) imaging and ex vivo biodistribution in murine xenografts of nontransfected PD-L1 expressing tumor cell lines, MDA-MB-231 (human breast carcinoma), and H2009 (human lung adenocarcinoma). It showed a favorable biodistribution profile with rapid renal clearance and low background (tumor-to-blood ratio = 26.6, 3 h p.i.). Conjugation of the Ga-68-TRAP moiety to WL12 successfully mitigated the nonspecific uptake of this peptide in organs, particularly the liver. This was demonstrated by comparing Ga-68-TRAP-WL12 with the archetypical Ga-68-DOTA-WL12, for which tumor-to-liver ratios of 1.4 and 0.5, respectively, were found. Although immunohistochemistry (IHC) revealed a low PD-L1 expression in MDA-MB-213 and H2009 xenografts that corresponds well to the clinical situation, PET showed high tumor uptakes (6.6 and 7.3% injected activity per gram of tissue (iA/g), respectively) for Ga-68-TRAP-WL12. Thus, this tracer has the potential for routine clinical PD-L1 PET imaging because it detects even very low PD-L1 expression densities with high sensitivity and may open an avenue to replace PD-L1 IHC of biopsies as the standard means to select potential responders for ICT.

Preclinical Evaluation and a Pilot Clinical Positron Emission Tomography Imaging Study of [68Ga]Ga-FAPI-FUSCC-II.

Fibroblast activation protein (FAP), a type II integral membrane serine protease, is a promising target for tumor diagnosis and therapy. OncoFAP has been recently discovered for PET imaging procedures for various solid malignancies. In this study, we presented the development of manual radiolabeling procedures for the preparation of OncoFAP-based radiopharmaceuticals for cancer imaging. A novel series of [68Ga/177Lu]Ga/Lu-FAPI-FUSCC-I/II were produced with high radiochemical yields. [68Ga]Ga-FAPI-FUSCC-I/II and [177Lu]Lu-FAPI-FUSCC-I/II were stable in phosphate-buffered saline, fetal bovine serum, and human serum for at least 3 h. In vitro cellular uptake and blocking experiments implied that they had specificity to FAP. Additionally, the low nanomolar IC50 values of FAPI-FUSCC-II indicated that it had a high target affinity to FAP. The in vivo biodistribution and blocking study in mice bearing HT-1080-FAP tumors showed that both exhibited specific tumor uptake. [68Ga]Ga-FAPI-FUSCC-II showed a higher tumor uptake and a higher tumor/nontarget ratio than [68Ga]Ga-FAPI-FUSCC-I and [68Ga]Ga-FAPI-04. The results of ex vivo biodistribution were in accordance with the biodistribution results. Clinical [68Ga]Ga-FAPI-FUSCC-II-PET/CT imaging further demonstrated its favorable biodistribution and kinetics with elevated and reliable uptake by primary tumors (maximum standardized uptake value (SUVmax), 12.17 ± 6.67) and distant metastases (SUVmax, 9.24 ± 4.28). In summary, [68Ga]Ga-FAPI-FUSCC-II displayed increased tumor uptake and retention compared to [68Ga]Ga-FAPI-04, giving it potential as a promising tracer for the diagnostic imaging of malignant tumors with positive FAP expression.

Location-Aware Encoding for Lesion Detection in 68Ga-DOTATATE Positron Emission Tomography Images.

Lesion detection with positron emission tomography (PET) imaging is critical for tumor staging, treatment planning, and advancing novel therapies to improve patient outcomes, especially for neuroendocrine tumors (NETs). Current lesion detection methods often require manual cropping of regions/volumes of interest (ROIs/VOIs) a priori, or rely on multi-stage, cascaded models, or use multi-modality imaging to detect lesions in PET images. This leads to significant inefficiency, high variability and/or potential accumulative errors in lesion quantification. To tackle this issue, we propose a novel single-stage lesion detection method using only PET images. We design and incorporate a new, plug-and-play codebook learning module into a U-Net-like neural network and promote lesion location-specific feature learning at multiple scales. We explicitly regularize the codebook learning with direct supervision at the network's multi-level hidden layers and enforce the network to learn multi-scale discriminative features with respect to predicting lesion positions. The network automatically combines the predictions from the codebook learning module and other layers via a learnable fusion layer. We evaluate the proposed method on a real-world clinical 68Ga-DOTATATE PET image dataset, and our method produces significantly better lesion detection performance than recent state-of-the-art approaches. We present a novel deep learning method for single-stage lesion detection in PET imaging data, with no ROI/VOI cropping in advance, no multi-stage modeling and no multi-modality data. This study provides a new perspective for effective and efficient lesion identification in PET, potentially accelerating novel therapeutic regimen development for NETs and ultimately improving patient outcomes including survival.

Abstract A068: Site-specific modification of nanobodies for ImmunoPET of Liver Cancer utilizing Self-labeling nanobody-tag pair (SLANT) technology

Abstract Visualization of tumor cells is critical in diagnosing, monitoring, and treating patients with cancer. Liver cancer has an 18 percent 5-year survival rate and is the third leading cause of cancer-related deaths worldwide. Hepatocellular carcinoma (HCC), the most common type of liver cancer, only takes up 18F-fluorodeoxyglucose (18F-FDG), the most commonly used clinical positron emission tomography (PET) tracer, 50% of the time. As such, developing an HCC-selective imaging agent could improve our ability for earlier diagnosis, post-treatment surveillance, and ability to increase patient longevity and quality of life. Single domain antibodies (sdAbs), or nanobodies (Nb), are attractive scaffolds for radiopharmaceutical development because of their high target affinity and specificity, ready availability from recombinant expression in bacteria, and small size for rapid clearance, better tumor penetration, and same-day imaging. However, conventional conjugation chemistry that exploits cysteines or lysines found on the native protein can abrogate the binding site of the smaller Nbs. We have previously used sortase A-based site-specific conjugation to address these challenges; however, improvements were still possible. Here, we developed a novel self-labeling Nb-tag pair (SLANT) that allows fast and specific covalent labeling of a Nb based on peptide epitope binding. We used SLANT to engineer a heterodimeric Nb construct consisting of a Nb (HN3) specific for Glypican-3, a heparin sulfate proteoglycan that is overexpressed in 75-90% of all HCCs and our self-labeling Nb. This design allows for site-specific labeling of the heterodimeric (GPC3-specific) construct in a manner that avoids modification of the tumor antigen-specific protomer. Labeling is achieved via exposure of the fusion protein to a crosslinking peptide functionalized with a chelator (DOTA). This methodology produces labeled protein at over 90 percent yields within minutes compared to less than 30 percent yields utilizing enzymatic methods. Biolayer interferometry and radioligand saturation studies were done to determine the binding affinity of the conjugate for GPC3 pre- and post-Indium-111 labeling. SPECT/CT and biodistribution studies were done at 1h, 3hrs, and 24hrs with HepG2 and HepG2-GPC3- xenografts to assess conjugate specificity and ability to accumulate in GPC3+ liver tumors. Our conjugate exhibited a nanomolar affinity for GPC3 in vitro (17-36 nM). SPECT/CT and biodistribution studies showed specificity for GPC3 in HepG2 liver cancer xenografts (~5.5% IA/g at 1-hour post-injection (p.i.)) and minimal uptake in HepG2 GPC3- tumors (~0.62% IA/g at 1-hour p.i.). Our results highlight the development of a new labeling paradigm applied to GPC3+ liver cancer. This approach can be used for same-day imaging and the rapid development of SPECT and PET imaging tracers for different cellular markers on cancer cells, addressing known tumor heterogeneity. Citation Format: Stanley Fayn, Chino C Cabalteja, Woonghee Lee, Divya Nambiar, Kwamena E Baidoo, Freddy E Escorcia, Ross W Cheloha. Site-specific modification of nanobodies for ImmunoPET of Liver Cancer utilizing Self-labeling nanobody-tag pair (SLANT) technology [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr A068.

Predictive value of metabolic parameters derived from preoperative 18F-FDG positron emission tomography/computed tomography for brain metastases in patients with surgically resected non-small cell lung cancer.

Brain metastases (BMs) are common complications in patients with non-small cell lung cancer (NSCLC). The purpose of this study was to investigate whether the metabolic parameters derived from preoperative 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) can predict BM development in patients with surgically resected NSCLC. We retrospectively reviewed 128 consecutive patients with stage I-IIIA NSCLC who underwent 18F-FDG PET/CT before curative surgery at The First Affiliated Hospital of Jinan University between November 2012 and October 2021. By drawing a volume of interest (VOI), the maximum standardized uptake values (SUVmax), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) of the primary tumor as well as the mean SUV (SUVmean) of the liver and arterial blood were measured. The tumor-to-liver SUV ratio (TLR) and tumor-to-blood SUV ratio (TBR) were also calculated. Receiver operating characteristic curve analysis was used to determine the best cut-off values for positron emission tomography (PET) parameters to predict BM-free survival, and Cox proportional hazards regression analysis was used to assess the predictive value of clinical variables and PET parameters. The median follow-up duration for survival patients was 23.4 months, and 15 patients (11.7%) experienced BM as the initial relapse site. The cumulative rates of BM over the course of 1, 2, and 5 years were 4.5%, 10.5%, and 17.5%, respectively. The optimal cut-off values for the prediction of BM-free survival were 7.7, 4.9, and 4.5 for SUVmax, TLR, and TBR, and 5.5 mL and 16.1 for MTV and TLG, respectively. In the Cox proportional hazards model, the risk of BM was significantly associated with TLR [hazard ratio (HR) =10.712; 95% confidence interval (CI): 2.958-38.801; P<0.001] and MTV (HR =3.150; 95% CI: 0.964-10.293; P=0.020) after adjusting for tumor stage, clinicopathological factors, and other PET parameters. Preoperative TLR and MTV of the primary tumor may be helpful in predicting BM development in patients with surgically resected NSCLC. Tumor metabolic parameters may potentially be used to stratify the risk of BM and determine individualized surveillance strategies.