检索结果-内蒙古大学图书馆

Generative adversarial networks in medical image reconstruction: A systematic literature review

Computers in Biology and Medicine 2025年 191卷 110094-110094页

作者： Hussain, Jabbar Båth, Magnus Ivarsson, Jonas Dept. of Applied IT University of Gothenburg Forskningsgången 6 417 56 Sweden Department of Medical Radiation Sciences University of Gothenburg Sweden

Purpose: Recent advancements in generative adversarial networks (GANs) have demonstrated substantial potential in medical image processing. Despite this progress, reconstructing images from incomplete data remains a challenge, impacting image quality. This systematic literature review explores the use of GANs in enhancing and reconstructing medical imaging data. Method: A document survey of computing literature was conducted using the ACM Digital Library to identify relevant articles from journals and conference proceedings using keyword combinations, such as "generative adversarial networks or generative adversarial network," "medical image or medical imaging," and "image reconstruction." Results: Across the reviewed articles, there were 122 datasets used in 175 instances, 89 top metrics employed 335 times, 10 different tasks with a total count of 173, 31 distinct organs featured in 119 instances, and 18 modalities utilized in 121 instances, collectively depicting significant utilization of GANs in medical imaging. The adaptability and efficacy of GANs were showcased across diverse medical tasks, organs, and modalities, utilizing top public as well as private/synthetic datasets for disease diagnosis, including the identification of conditions like cancer in different anatomical regions. The study emphasized GAN's increasing integration and adaptability in diverse radiology modalities, showcasing their transformative impact on diagnostic techniques, including cross-modality tasks. The intricate interplay between network size, batch size, and loss function refinement significantly impacts GAN's performance, although challenges in training persist. Conclusions: The study underscores GANs as dynamic tools shaping medical imaging, contributing significantly to image quality, training methodologies, and overall medical advancements, positioning them as substantial components driving medical advancements. © 2025 The Authors

关键词： Medical image processing

来源：评论

学校读者我要写书评

暂无评论

image reconstruction from radially incomplete spherical Radon data

引用

EUROPEAN JOURNAL OF APPLIED MATHEMATICS 2018年第3期29卷 470-493页

作者： Ambartsoumian, Gaik Gouia-Zarrad, Rim Krishnan, Venkateswaran P. Roy, Souvik Univ Texas Arlington Dept Math Arlington TX 76019 USA Amer Univ Sharjah Dept Math & Stat Sharjah U Arab Emirates Tata Inst Fundamental Res Ctr Applicable Math Bangalore Karnataka India Univ Wurzburg Dept Math Wurzburg Germany

We study inversion of the spherical Radon transform with centres on a sphere (the data acquisition set). Such inversions are essential in various image reconstruction problems arising in medical, radar and sonar imaging. In the case of radially incomplete data, we show that the spherical Radon transform can be uniquely inverted recovering the image function in spherical shells. Our result is valid when the support of the image function is inside the data acquisition sphere, outside that sphere, as well as on both sides of the sphere. Furthermore, in addition to the uniqueness result, our method of proof provides reconstruction formulas for all those cases. We present a robust computational algorithm and demonstrate its accuracy and efficiency on several numerical examples.

关键词： spherical Radon transform medical imaging spherical harmonics Volterra integral equations truncated singular value decomposition

来源：评论

学校读者我要写书评

暂无评论

Bayesian image and pattern reconstruction from incomplete and noisy data

引用

Pattern Recognition and image Analysis 2012年第1期22卷 99-107页

作者： Vengrinovich, V.L. Institute of Applied Physics National Academy of Sciences Minsk Belarus

A Bayesian iterative method can be the basis for a wide range of technologies in the field of pattern recognition and image reconstruction. It involves finding the most probable solutions for images or patterns, if functionals describing the likelihood function and a priori information, respectively, are already known. The article describes the basic principles and recent advances in the development of BIM and its applications in various fields, mainly in tomography and restoration of functions from incomplete and noisy data. © 2012 Pleiades Publishing, Ltd.

关键词： Restoration

来源：评论

学校读者我要写书评

暂无评论

A Trust-Guided Approach to MR image reconstruction with Side Information

arXiv

引用

arXiv 2025年

作者： Atalık, Arda Chopra, Sumit Sodickson, Daniel K. The Bernard and Irene Schwartz Center for Biomedical Imaging Department of Radiology NYU Grossman School of Medicine United States The Courant Institute of Mathematical Sciences The Bernard and Irene Schwartz Center for Biomedical Imaging Department of Radiology NYU Grossman School of Medicine United States The Bernard and Irene Schwartz Center for Biomedical Imaging Department of Radiology NYU Grossman School of Medicine United States

Reducing MRI scan times can improve patient care and lower healthcare costs. Many acceleration methods are designed to reconstruct diagnostic-quality images from limited sets of acquired k-space data. This task can be framed as a linear inverse problem (LIP), where, as a result of undersampling, the forward operator may become rank-deficient or exhibit small singular values. This results in ambiguities in reconstruction, in which multiple generally incorrect or non-diagnostic images can map to the same acquired data. To address such ambiguities, it is crucial to incorporate prior knowledge, for example in the form of regularization. Another form of prior knowledge less commonly used in medical imaging is contextual side information garnered from other sources than the current acquisition. Here, we propose the Trust-Guided Variational Network (TGVN), a novel end-to-end deep learning framework that effectively integrates side information into LIPs. TGVN eliminates undesirable solutions from the ambiguous space of the forward operator while remaining faithful to the acquired data. We demonstrate its effectiveness in multi-coil, multi-contrast MR image reconstruction, where incomplete or low-quality measurements from one contrast are used as side information to reconstruct high-quality images of another contrast from heavily under-sampled data. Our method is robust across different contrasts, anatomies, and field strengths. Compared to baselines that also utilize side information, TGVN achieves superior image quality at challenging under-sampling levels, drastically speeding up acquisition while minimizing hallucinations. Our approach is also versatile enough to incorporate many different types of side information (including previous scans or even text) into any LIP. © 2025, CC BY-NC-ND.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

An unsupervised method for MRI recovery: Deep image prior with structured sparsity

arXiv

引用

arXiv 2025年

作者： Sultan, Muhammad Ahmad Chen, Chong Liu, Yingmin Gil, Katarzyna Zareba, Karolina Ahmad, Rizwan The Ohio State University ColumbusOH43210 United States Davis Heart and Lung Research Institute The Ohio State University Wexner Medical Center ColumbusOH43210 United States Division of Cardiovascular Medicine The Ohio State University Wexner Medical Center ColumbusOH43210 United States

Objective: To propose and validate an unsupervised MRI reconstruction method that does not require fully sampled k-space data. Materials and Methods: The proposed method, deep image prior with structured sparsity (DISCUS), extends the deep image prior (DIP) by introducing group sparsity to frame-specific code vectors, enabling the discovery of a low-dimensional manifold for capturing temporal variations. DISCUS was validated using four studies: (I) simulation of a dynamic Shepp-Logan phantom to demonstrate its manifold discovery capabilities, (ii) comparison with compressed sensing and DIP-based methods using simulated single-shot late gadolinium enhancement (LGE) image series from six distinct digital cardiac phantoms in terms of normalized mean square error (NMSE) and structural similarity index measure (SSIM), (iiI) evaluation on retrospectively undersampled single-shot LGE data from eight patients, and (IV) evaluation on prospectively undersampled single-shot LGE data from eight patients, assessed via blind scoring from two expert readers. Results: DISCUS outperformed competing methods, demonstrating superior reconstruction quality in terms of NMSE and SSIM (Studies I-iiI) and expert reader scoring (Study IV). Discussion: An unsupervised image reconstruction method is presented and validated on simulated and measured data. These developments can benefit applications where acquiring fully sampled data is challenging. © 2025, CC BY-NC-SA.

关键词： Phantoms

来源：评论

学校读者我要写书评

暂无评论

ONE-STEP image-reconstruction from incomplete data IN COMPUTER-TOMOGRAPHY

引用

IEEE TRANSACTIONS ON MEDICAL IMAGING 1989年第1期8卷 16-31页

作者： PENG, H STARK, H RENSSELAER POLYTECH INST DEPT ELECT COMP & SYST ENGNTROYNY 12180

The authors develop a projection operator that simultaneously projects onto the set of all functions satisfying raysum constraints in parallel-beam CT. The projector can be directly extended to the fan-beam case through the process of rebinning. The projector generates a solution that is closest to the initial estimate among all the functions that are consistent with the available raysum data. It realizes the ART (algebraic reconstruction technique) in one step. The projector furnishes the one-step projection reconstruction (OSPR) for any arbitrary configurations of missing data. Because the projection is one-step, there can be a significant reduction in the number of online computations and memory requirements, especially when the missing data exhibit some pattern within a view or between views.< >

关键词： image reconstruction Computed tomography Subspace constraints Iterative algorithms Concurrent computing Systems engineering and theory Inspection X-ray imaging Bayesian methods

来源：评论

学校读者我要写书评

暂无评论

RIS: Regularized imaging spectroscopy for STIX on board Solar Orbiter

引用

ASTRONOMY & ASTROPHYSICS 2025年 694卷

作者： Volpara, Anna Lupoli, Alessandro Filbir, Frank Perracchione, Emma Massone, Anna Maria Piana, Michele Univ Genoa Dipartimento Matemat MIDA Via Dodecaneso 35 I-16146 Genoa Italy Tech Univ Munich TUM Sch Computat Informat & Technol Dept Math Boltzmannstr 3 D-85748 Garching B Munchen Germany Helmholtz Munich Inst Biol & Med Imaging Bioengn Ctr Ingolstadter Landstr 1 D-85764 Neuherberg Germany Politecn Torino Dipartimento Sci Matemat Giuseppe Luigi Lagrange Corso Duca Abruzzi 24 I-10129 Turin Italy Ist Nazl Astrofis Osservatorio Astrofisico Torino Via Osservatorio 20 I-10025 Pino Torinese Italy

Context. The generation of spatially resolved count spectra and of cubes of count maps at different energies via imaging spectroscopy is one of the main goals of solar hard X-ray missions based on Fourier imaging. Thus, so far, for these telescopes, this goal has been realized via the generation of either count maps that are independently reconstructed in the different energy channels or electron flux maps reconstructed via a deconvolution of the approximate forms for the bremsstrahlung cross-section. Aims. Our aim is to introduce a regularized imaging spectroscopy method (RIS), whereby the regularization implemented in the count space imposes a smoothing constraint across contiguous energy channels, without the need for computing any deconvolution of the bremsstrahlung effect. Methods. STIX records imaging data, while computing the visibilities in the spatial frequency domain. Our RIS is a sequential scheme in which part of the information coded in the image is reconstructed at a specific energy channel and transferred to the reconstruction process at a contiguous channel via a visibility interpolation computed by means of variably scaled kernels (VSKs). Results. In the case of STIX visibilities recorded during the November 11, 2022 flaring event, we show that RIS is able to generate hard X-ray maps, whose morphology is seen to smoothly evolve from one energy channel to the contiguous one;accordingly, from these maps, it is possible to infer spatially resolved count spectra characterized by a notable numerical stability. We also show that the performance of this approach is robust with respect to both the image reconstruction method and the count energy channel utilized to trigger the sequential process. Conclusions. We conclude that RIS is not only an appropriate, but also an effective and necessary approach to constructing image cubes from STIX visibilities that are characterized by smooth behavior across count energies. Thus, it allows for the generation of nume

关键词： methods: data analysis techniques: imaging spectroscopy telescopes Sun: flares Sun: X-rays gamma rays

来源：评论

学校读者我要写书评

暂无评论

Deep learning filtered back projection for photoacoustic tomography

Deep learning filtered back projection for photoacoustic tom...

引用

Photons Plus Ultrasound: Imaging and Sensing 2025

作者： Tian, Chao Sun, Junjun Liu, Songde University of Science and Technology of China Anhui Hefei230026 China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Anhui Hefei230088 China School of Artificial Intelligence Anhui University Anhui Hefei230601 China

ISBN: (纸本)9781510683860

image reconstruction in photoacoustic tomography (PAT) faces significant challenges due to incomplete projection data. To address this problem, we propose a physics-driven, deep learning-based filtered back-projection algorithm (dFBP). The dFBP framework is systematically designed based on the principles of the traditional analytical filtered back-projection algorithm and consists of a filtering module, a back-projection module, and a fusion module arranged in a cascaded architecture. Numerical simulations show that dFBP-based PAT accurately reconstructs high-quality, artifact-suppressed images from sparse, limited, and acoustically heterogeneous projection data. © 2025 SPIE.

关键词： image restoration reconstruction algorithms Deep learning image quality Photoacoustic tomography

来源：评论

学校读者我要写书评

暂无评论

Canonical Shape reconstruction with SE(3) Equivariance Learning for Weakly-Supervised Object Pose Estimation

引用

IEEE Transactions on Circuits and Systems for Video Technology 2025年

作者： Zhou, Jun Chen, Kai Wei, Mingqiang Zhang, Xiao-Ping Dou, Qi Qin, Jing Hong Kong Polytechnic University Center of Smart Health School of Nursing Hong Kong Hong Kong Chinese University of Hong Kong Department of Computer Science and Engineering Hong Kong Hong Kong Nanjing University of Aeronautics and Astronautics School of Computer Science and Technology Nanjing China Tsinghua University Shenzhen Ubiquitous Data Enabling Key Lab Shenzhen International Graduate School Shenzhen China

6D object pose estimation from a single RGB-D image is a fundamental problem in computer vision and robot manipulation. Despite recent advancements, existing methods still suffer several limitations. First of all, the object shape representation extracted from the depth map is often less expressive because the object point cloud parsed from the depth map is highly incomplete due to the object self-occlusion and noisy due to the sensor artifacts. This shape representation issue further intensifies when lacking sufficient labeled data for model training, which unfortunately is another typical problem for object pose estimation considering the heavy annotation cost for real-world pose labeling. In this study, we propose to tackle the above issues in a unified way. First, we enhance the object shape representation from the partial point cloud with a novel canonical shape reconstruction module, in which an implicit canonical frame is established by incorporating the SE(3) equivariance, achieving implicit feature alignment of the partial point cloud inputs, leading to robust shape recovery. Second, based on the enhanced object representation, we further utilize the decanonicalized and pose-dependent completed object shape as the training signal, and develop a novel weakly-supervised learning framework to leverage both labeled synthetic data and unlabeled real data to train the pose estimation model in a label-efficient way. Extensive experiments on three widely used benchmarks demonstrate the effectiveness, and superiority of our framework over state-of-the-art methods. © 1991-2012 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Two-Phase Pressure Drops Analysis in a Chevron-Type Corrugated Channel Using image Segmentation for Void Fraction Estimation 41

Two-Phase Pressure Drops Analysis in a Chevron-Type Corrugat...

引用

41st UIT International Heat Transfer Conference

作者： Passoni, S. Ferrario, A. Mereu, R. Inzoli, F. Politecnico di Milano Department of Energy via Lambruschini 4A Milano20156 Italy

This study examines the pressure drops and flow patterns in a Chevron-type Compact Plate Heat Exchanger (CPHE) with air-water two-phase mixture under adiabatic conditions. A transparent replica channel, made from epoxy resin, was used to mimic the heat exchanger's fluid dynamics and allow for optical access. The channel featured a 63° corrugation angle, 2.5 mm depth, and 8.9 mm pitch, yielding a global enlargement factor of 1.17. A particular inlet section was designed to ensure even phase mixing and distribution. Operating conditions varied from 6 to 365 kg/m2s for water and 0.02 to 5 kg/m2s for air, expanding the range of previous studies. High-framerate back-illuminated visualizations provided detailed observations of the flow patterns. Moreover, frame-by-frame void fraction was estimated via HSV video analysis adopting a convolutional neural network (U-Net) to produce binary masks and estimate gas bubble volumes and via a volume-reconstruction algorithm exploiting the pseudo-2D shape of the channel. A novel correlation for Darcy's friction factor based on Reynolds number was developed for single-phase flow, complemented by an analysis of two-phase pressure drops using a two-phase multiplier correlated with the Lockhart-Martinelli parameter. The incorporation of void fraction data significantly improved their modelling accuracy. © 2025 Institute of Physics Publishing. All rights reserved.

关键词： Reynolds number

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：