IEEE Transactions on Radiation and Plasma Medical Sciences最新文献

{"title":"Development and Initial Evaluation of 3D-Printed High Resolution Brain Phantom for PET","authors":"Cynthia Lo;Ekaterina Shanina;Paul Gravel;Yifan Hu;Steven A. Lucero;Sean Martins;Tim Mulnix;Kathryn Fontaine;Seyed Faraz Nejati;Xishan Sun;Hongdi Li;Jinyi Qi;Simon R. Cherry;Richard E. Carson;Takuya Toyonaga","doi":"10.1109/TRPMS.2025.3638588","DOIUrl":"10.1109/TRPMS.2025.3638588","url":null,"abstract":"Phantoms are essential for assessing positron emission tomography (PET) system performance, but most existing brain phantoms do not provide the anatomical fidelity required for the next-generation high-resolution PET systems. To address this limitation, we developed HYDRA-OL (HYDrodynamically filled Realistic Anatomy of the Occipital Lobe) phantoms using high-resolution 3-D printing based on the BigBrain atlas. In this study, two phantom types were created, HYDRA-OLG (gray matter (GM) only) and HYDRA-OLGW (gray and white matter (WM)), enabling contrast modulation between GM and WM. Precise anatomical fidelity and repositioning accuracy were confirmed via CT. A custom closed-loop filling apparatus ensured efficient and reproducible phantom preparation. Residual air in the phantom after 15 min was estimated to be less than 0.01% of the fillable space. To compare spatial resolution between scanners, the HYDRA-OLG phantom was imaged on four different PET systems (NeuroEXPLORER (NX), high resolution research tomograph (HRRT), Vision, and Focus220), with NX and Focus showing superior spatial resolution compared to the other two systems. To generate variable GM:WM contrast, list mode datasets obtained on the NX for both HYDRA-OLG and HYDRA-OLGW phantoms were down-sampled and merged to form one list mode data. Reconstructed images were successfully produced with GM:WM contrast ratio of 4:1, 2:1, and 4:3. Overall, the HYDRA phantom platform is a versatile tool for accurately representing human brain anatomy. It is likely to be more useful than previous phantoms for evaluating image resolution and optimizing reconstruction parameters for next-generation brain PET systems.","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 4","pages":"603-610"},"PeriodicalIF":3.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145828742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ResPF: Residual Poisson Flow Generative Model for Efficient and Physically Consistent Sparse-View CT Reconstruction","authors":"Changsheng Fang;Yongtong Liu;Bahareh Morovati;Shuo Han;Yu Shi;Li Zhou;Shuyi Fan;Hengyong Yu","doi":"10.1109/TRPMS.2025.3615836","DOIUrl":"https://doi.org/10.1109/TRPMS.2025.3615836","url":null,"abstract":"Sparse-view computed tomography (CT) reduces radiation dose but leads to an ill-posed inverse problem that challenges accurate reconstruction. While deep learning and diffusion-based methods have shown promising results, they often lack physical consistency or suffer from high-computational cost due to iterative sampling from random noise. Recent advances in generative modeling, especially poisson flow generative models (PFGMs), offer high-fidelity synthesis by modeling the full data distribution via deterministic ordinary differential equation (ODE) trajectories. In this work, we propose residual poisson flow (ResPF), the first application of PFGM++ to sparse-view CT reconstruction. ResPF introduces the four key innovations: 1) a conditional PFGM++ trained on paired sparse/full-view data; 2) a hijacking strategy that skips early sampling steps to accelerate convergence; 3) a data-consistency module embedded at each iteration to enforce fidelity to measured projections; and 4) to preserve the stability of the generative path, we further propose a residual fusion mechanism to combine generative outputs with physics-consistent updates. Extensive experiments on synthetic and clinical datasets demonstrate that ResPF achieves state-of-the-art performance, reaching 0.964 structural similarity (SSIM) and 39.8 dB peak signal-to-noise ratio (PSNR) on 63-view reconstructions (from 1000-view groundtruth), significantly outperforming existing baselines. Moreover, ResPF reduces sampling time to just 1.5 s per image that is over <inline-formula> <tex-math>$15{times }$ </tex-math></inline-formula> faster than conventional diffusion-based methods, highlighting its practicality for real-world applications.","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 4","pages":"520-534"},"PeriodicalIF":3.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147588230","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dynamic PET Image Reconstruction Using Kalman Inspired Network","authors":"Mengrui Chen;Hengjia Ran;Yiming Wan;Ao Ran;Min Guo;Huafeng Liu","doi":"10.1109/TRPMS.2025.3602938","DOIUrl":"https://doi.org/10.1109/TRPMS.2025.3602938","url":null,"abstract":"Positron Emission Tomography (PET) facilitates the visualization of the distribution of radioactive isotope-labeled compounds within biological organisms. Existing reconstruction algorithms, including iterative and deep learning methods that depend on the system matrix, often suffer from inaccuracies in the system matrix, resulting in artifacts or blurring in the reconstructed images. Inspired by Kalman Filtering, we propose a data-driven deep learning framework for PET image reconstruction, named the Kalman Inspired Network (KIN). The KIN framework divides the reconstruction problem into two phases: 1) state prediction and 2) state update, and consists of three core components: 1) prediction net; 2) projection net; and 3) Kalman Gain Net. By adopting a data-driven approach, KIN circumvents the reliance on the system matrix, thereby overcoming the limitations imposed by noise prior knowledge and the inversion of large-dimensional matrices typically required in traditional Kalman Filtering algorithms. We evaluated the proposed KIN network using simulated phantom and experimental rat datasets, benchmarking it against traditional algorithms as well as other deep-learning-based methods. The results demonstrate that KIN enhances the quality of dynamic PET scans, both in terms of image quality and quantitative indices, underscoring its potential for dynamic imaging applications that require extremely short frames and are typically challenged by high noise levels.","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 4","pages":"492-503"},"PeriodicalIF":3.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147588245","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DOI Self-Calibration Method Based on Intrinsic Radiation for Positron Emission Tomography","authors":"Xin Yu;Haoyu Zou;Han Liu;Zhijun Zhao;Da Liang;Xiaoyin Zhang;Huiping Zhao;Jinyong Tao;Wenjing Zhu;Yibin Zhang;Jiayang Zeng;Xiaoying Tang;Jianfeng Xu;Qiyu Peng","doi":"10.1109/TRPMS.2025.3602383","DOIUrl":"https://doi.org/10.1109/TRPMS.2025.3602383","url":null,"abstract":"depth of interaction (DOI) decoding can reduce localization errors in the line-of-response (LOR), thereby improving positron emission tomography (PET) image quality, especially in high-resolution PET, long-axial PET, and small-aperture PET systems. Current PET detectors are capable of achieving high-resolution DOI decoding, but most methods rely on collimation calibration for each detector. Calibrating all detectors is labor-intensive and complicates subsequent system maintenance. We aim to achieve DOI collimation for PET detectors based on the intrinsic radiation of 176Lu, which allows us to omit the complex step of calibrating DOI for all detectors. This approach allows DOI calibration to be performed automatically during each system operation, simplifying system maintenance. The pixelated LYSO detector in this study consists of LYSO pixels with dimensions of <inline-formula> <tex-math>$ 2 times 2 times 20 , text {mm}^{3} $ </tex-math></inline-formula>. The detector is arranged in an <inline-formula> <tex-math>$8times 10$ </tex-math></inline-formula> array configuration. We obtained the relationship between intrinsic radiation and DOI through Monte Carlo simulations, and the relationship between the multi pixel photon counter (MPPC) signal ratio (the relative proportion of light detected by different MPPCs from the same crystal, which changes with DOI) and intrinsic radiation through experiments. This allowed us to establish a conversion function between DOI and the MPPC signal ratio. We verified the DOI resolution through collimation testing and demonstrated the improvement in image quality of the DOI self-calibration based on intrinsic radiation method using the Derenzo phantom experiment. We validated the DOI resolution of our method through mechanical collimation experiments at depths of 4, 10, and 16 mm, achieving an average DOI resolution full width at half maximum (FWHM) of 5.2 mm and a mean absolute error (MAE) of 3.6 mm. The images reconstructed with DOI correction obtained through self-collimation demonstrated significantly improved signal-to-noise ratio and spatial resolution compared to images without DOI correction. Specifically, the spatial resolution at an offset of 45 mm improved from 3.0 to 2.1 mm. The DOI self-calibration method based on intrinsic radiation provides a practical and efficient solution for calibrating PET detectors without the need to add complex hardware. It should be noted that this method is only applicable to PET systems with detectors employing LSO or LYSO, and therefore this method is based on spontaneous emission of <inline-formula> <tex-math>${}^{176}text {Lu}$ </tex-math></inline-formula>. This method has the potential to improve the image quality of PET systems and can be readily integrated into existing detector designs.","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 4","pages":"574-581"},"PeriodicalIF":3.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147588213","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Light Propagation Modeling in a Mie-Scattering Dominated Scintillator","authors":"A. Dahmane;M. Bongrand;C. Bourgeois;D. Breton;M. Briere;A. Cabrera;V. Chaumat;R. Gazzini;D. Giovagnoli;F. Haddad;A. Hourlier;G. Hull;M. Hussein;P. Lanièce;F. Lefevre;P. Loaiza;J. Maalmi;Y. Mellak;T. Merlin;R. Mastrippolito;C. Marquet;L. Ménard;D. Navas-Nicolàs;P. Pillot;L. Simard;D. Stocco;M.-A. Verdier;D. Visvikis;F. Yermia;D. Brasse","doi":"10.1109/TRPMS.2025.3614151","DOIUrl":"https://doi.org/10.1109/TRPMS.2025.3614151","url":null,"abstract":"We present an analytical approach to model light propagation in a highly scattering, opaque organic scintillator, as used in the novel positron emission tomography detector concept LiquidO. In this detection scheme, annihilation photons undergo multiple Compton scatterings, each producing a localized light emission. The scintillating medium being opaque, the scintillation light is confined near its emission point, mainly by Mie scattering. Each annihilation photon interaction is characterized by a succession of light balls, collected by a lattice of optical fibers that are read out at both ends. Simulating optical photons using Monte-Carlo (MC) methods, as implemented in <sc>Geant4</small> and <sc>Gate</small>, requires substantial computational resources and introduces statistical noise. To address this, we developed an analytical model (AM) to predict the spatial and temporal probability density function of optical photons in a scintillating, scattering and absorbing medium. The model combines isotropic random walk dynamics with a 2-D polynomial parameterization to account for anisotropic scattering, by correcting the transport parameters for a double-lobe Henyey–Greenstein phase function with opposite asymmetry factors and a tunable forward-to-backward lobe ratio. Validated against MC simulations, the model predicts spatial and temporal photon distributions with an average relative deviation of 1.60% for the mean radius per time slice <inline-formula> <tex-math>$r(t)$ </tex-math></inline-formula> and 0.42% for the mean time per radial slice <inline-formula> <tex-math>$t(r)$ </tex-math></inline-formula>, while being between 289 and 967 times faster than MC simulations. Its validity spans optical parameters with scattering mean free path of 2–5 mm, asymmetry factor between 0.3–0.7, and forward-to-backward scattering ratio of 0.25–1. The proposed AM can be used to speed up optical simulations, with applications in detector design, event reconstruction, optimization of medium optical properties, and modeling of highly scattering media.","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 4","pages":"558-565"},"PeriodicalIF":3.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147588258","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spatial Dosimetry Verification of Conformal Beams in Proton Radiotherapy by Spectral-Sensitive Imaging","authors":"Carlos Granja;Jan Swakon;Pawel Olko","doi":"10.1109/TRPMS.2025.3603137","DOIUrl":"https://doi.org/10.1109/TRPMS.2025.3603137","url":null,"abstract":"Particle radiotherapy with conformal beams aims to tailor the delivered radiation and dose with spatial customization. For treatments using beam shaping collimators it is desired to verify the delivered doses and examine the characteristics of the produced fields with imaging response. For this purpose, we use the semiconductor pixel detector Timepix which provides time-resolved spectral and imaging response with high-spatial resolution and particle-type resolving power. Precise measurements are performed in-field at the isocenter on selected regions (<inline-formula> <tex-math>$leq $ </tex-math></inline-formula> 14 mm) of delivered beams of low intensity (<inline-formula> <tex-math>$leq $ </tex-math></inline-formula> pA). Detailed information in wide dynamic range is produced in the form of field composition (particle-type species), spectral (deposited energy, linear-energy-transfer (LET) spectra) and spatial dose distribution at the sub-mm scale (<inline-formula> <tex-math>$approx ~200$ </tex-math></inline-formula>-<inline-formula> <tex-math>$mu $ </tex-math></inline-formula>m spatial resolution). Results are presented for measurements of two conformal beam collimators used for eye treatment with 55 MeV protons. The produced 2-D maps and 1-D profile distributions can be used for experimental verification of spatially fractionated radiation therapy (grid therapy), and in particle therapy for verification of Monte Carlo transport calculation, including LET distribution at the target isocenter position.","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 4","pages":"611-618"},"PeriodicalIF":3.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147588269","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Feature Space-Guided Denoising of Noisy 4-D Data: Applications to Dynamic PET Imaging and Dual-Calibrated Functional MRI","authors":"Connor W. J. Bevington;Ju-Chieh Cheng;Rebecca J. Williams;Jing Zhang;Wen-Ming Luh;G. Bruce Pike;Vesna Sossi","doi":"10.1109/TRPMS.2025.3608506","DOIUrl":"https://doi.org/10.1109/TRPMS.2025.3608506","url":null,"abstract":"4-D neuroimaging methods, including dynamic positron emission tomography (PET) and functional MRI, capture the spatiotemporal behavior of physiological processes. Subsequent analysis, such as kinetic modeling of dynamic PET data, can provide parametric images related to unique aspects of physiology. However, 4-D data is often exceptionally noisy, thus requiring post-processing denoising for reliable quantitative results. Several proposed 4-D denoising algorithms reduce noise via signal averaging of physiologically similar voxels, but often have the tradeoff of reduced accuracy. Furthermore, many do not fully exploit nonlocal voxel similarity, due to computational constraints and/or a suboptimal heuristic for identifying physiological similarity. In this work, we propose a denoising framework that uses a low-dimensional feature space representation of the data to identify similar voxels. This permits targeted nonlocal denoising to produce an initial denoised product, which is then passed to the HighlY constrained backPRojection (HYPR) algorithm to ensure data consistency. We optimize the feature space for data from different PET tracers and demonstrate cross-modality applicability, using dual-calibrated fMRI as a proof-of-concept. Additionally, we show our proposed method is superior to comparable denoising algorithms in terms of quantitative accuracy and precision of parametric images computed from the denoised data—thus demonstrating improvements relevant for research and clinical applications.","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 4","pages":"504-519"},"PeriodicalIF":3.5,"publicationDate":"2026-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147588287","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IEEE Transactions on Radiation and Plasma Medical Sciences Information for Authors","authors":"","doi":"10.1109/TRPMS.2026.3663808","DOIUrl":"https://doi.org/10.1109/TRPMS.2026.3663808","url":null,"abstract":"","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 3","pages":"C3-C3"},"PeriodicalIF":3.5,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11421172","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147352560","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IEEE Transactions on Radiation and Plasma Medical Sciences Publication Information","authors":"","doi":"10.1109/TRPMS.2026.3663806","DOIUrl":"https://doi.org/10.1109/TRPMS.2026.3663806","url":null,"abstract":"","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 3","pages":"C2-C2"},"PeriodicalIF":3.5,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11421173","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147352565","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"IEEE DataPort","authors":"","doi":"10.1109/TRPMS.2026.3663800","DOIUrl":"https://doi.org/10.1109/TRPMS.2026.3663800","url":null,"abstract":"","PeriodicalId":46807,"journal":{"name":"IEEE Transactions on Radiation and Plasma Medical Sciences","volume":"10 3","pages":"472-472"},"PeriodicalIF":3.5,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11421171","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147352537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}