检索结果-内蒙古大学图书馆

Cyclical Learning Rate Optimization on deep Learning Model for Brain Tumor Segmentation

IEEE ACCESS 2023年 11卷 119802-119810页

作者： Fajar, Aziz Sarno, Riyanarto Fatichah, Chastine Susilo, Rahadian Indarto Pangestu, Gusti Inst Teknol Sepuluh Nopember Fac Intelligent Elect & Informat Technol Dept Informat Surabaya 60111 Indonesia Univ Airlangga Fac Adv Technol & Multidiscipline Dept Data Sci Technol Surabaya 60115 Indonesia Univ Airlangga Fac Med Dept Neurosurg Surabaya Indonesia Bina Nusantara Univ Sch Comp Sci Jakarta Indonesia

In recent years, deep learning has found widespread applications in tasks such as segmentation and classification. Fine-tuning hyperparameters is crucial to improve performance, with the learning rate being a key parameter. Various methods, including adaptive learning rates, learning rate scheduling, and cyclical learning rates, have been used to optimize learning rates. Cyclical learning rates offer significant benefits with minimal computational cost, as seen in previous research. This study introduces a novel approach to tuning the cyclical learning rate, which incorporates the exponential moving average. These methods are applied to the BraTS 2021 dataset for segmentation tasks, resulting in superior performance compared to the previous approach. Our proposed method reduces the epochs required to reach convergence by 19 and 54 epochs for U-Net and dense U-net, respectively. For Res U-net, the epoch needed to convergence is 10 epochs more. However, the proposed method produces lower loss values with 0.707, 0.657, and 0.665 compared to the previous method with 0.712, 0.685, and 0.725 for U-net, Res U-net, and dense U-net, respectively.

关键词： 3d image data deep learning exponential moving average information and communication technology learning rate scientific research

来源：评论

学校读者我要写书评

暂无评论

Stochastic 3d Modeling of Three-Phase Microstructures for Predicting Transport Properties: A Case Study

引用

TRANSPORT IN POROUS MEdIA 2019年第1期128卷 179-200页

作者： Neumann, M. Abdallah, B. Holzer, L. Willot, F. Schmidt, V. Ulm Univ Inst Stochast Helmholtzstr 18 D-89069 Ulm Germany PSL Res Univ Ctr Math Morphol MINES ParisTech Rue St Honore 35 F-77300 Fontainebleau France ZHAW Winterthur Inst Computat Phys Wildbachstr 21 CH-8400 Winterthur Switzerland

We compare two conceptually different stochastic microstructure models, i.e., a graph-based model and a pluri-Gaussian model, that have been introduced to model the transport properties of three-phase microstructures occurring, e.g., in solid oxide fuel cell electrodes. Besides comparing both models, we present new results regarding the relationship between model parameters and certain microstructure characteristics. In particular, an analytical expression is obtained for the expected length of triple phase boundary per unit volume in the pluri-Gaussian model. As a case study, we consider 3d image data which show a representative cutout of a solid oxide fuel cell anode obtained by FIB-SEM tomography. The two models are fitted to image data and compared in terms of morphological characteristics (like mean geodesic tortuosity and constrictivity) as well as in terms of effective transport properties. The Stokes flow in the pore phase and effective conductivities in the solid phases are computed numerically for realizations of the two models as well as for the 3d image data using Fourier methods. The local and effective physical responses of the model realizations are compared to those obtained from 3d image data. Finally, we assess the accuracy of the two methods to predict permeability as well as electronic and ionic conductivities of the anode.

关键词： Stochastic microstructure modeling Effective conductivity Permeability Solid oxide fuel cells 3d image data

来源：评论

学校读者我要写书评

暂无评论

Improved α-GAN architecture for generating 3d connected volumes with an application to radiosurgery treatment planning

引用

APPLIEd INTELLIGENCE 2023年第18期53卷 21050-21076页

作者： Jafari, Sanaz Mohammad Cevik, Mucahit Basar, Ayse Toronto Metropolitan Univ 44 Gerrard St E Toronto ON M5B 1G3 Canada

Generative Adversarial Networks (GANs) have gained significant attention in several computer vision tasks for generating high-quality synthetic data. Various medical applications including diagnostic imaging and radiation therapy can benefit greatly from synthetic data generation due to data scarcity in the domain. However, medical image data is typically kept in 3d space, and generative models suffer from the curse of dimensionality issues in generating such synthetic data. In this paper, we investigate the potential of GANs for generating connected 3d volumes. We propose an improved version of 3d alpha-GAN by incorporating various architectural enhancements. On a synthetic dataset of connected 3d spheres and ellipsoids, our model can generate fully connected 3d shapes with similar geometrical characteristics to that of training data. We also show that our 3d GAN model can successfully generate high-quality 3d tumor volumes and associated treatment specifications (e.g., isocenter locations). Similar moment invariants to the training data as well as fully connected 3d shapes confirm that improved 3d alpha-GAN implicitly learns the training data distribution, and generates realistic-looking samples. The capability of improved 3d alpha-GAN makes it a valuable source for generating synthetic medical image data that can help future research in this domain.

关键词： Generative adversarial networks 3d image data image synthesis Connected components Radiation therapy

来源：评论

学校读者我要写书评

暂无评论

Optimal designs for three-dimensional shape analysis with spherical harmonic descriptors

引用

ANNALS OF STATISTICS 2005年第6期33卷 2758-2788页

作者： dette, H Melas, VB Pepelyshev, A Ruhr Univ Bochum Fak Mathemat D-44780 Bochum Germany St Petersburg State Univ Dept Math St Petersburg 198904 Russia

We determine optimal designs for some regression models which are frequently used for describing three-dimensional shapes. These models are based on a Fourier expansion of a function defined on the unit sphere in terms of spherical harmonic basis functions. In particular, it is demonstrated that the uniform distribution on the sphere is optimal with respect to all Phi(p) criteria proposed by Kiefer in 1974 and also optimal with respect to a criterion which maximizes a p mean of the r smallest eigenvalues of the variance-covariance matrix. This criterion is related to principal component analysis, which is the common tool for analyzing this type of image data. Moreover, discrete designs on the sphere are derived, which yield the same information matrix in the spherical harmonic regression model as the uniform distribution and are therefore directly implementable in practice. It is demonstrated that the new designs are substantially more efficient than the commonly used designs in three-dimensional shape analysis.

关键词： shape analysis spherical harmonic descriptors optimal designs quadrature formulas principal component analysis 3d image data

来源：评论

学校读者我要写书评

暂无评论

CELL NUCLEI SEGMENTATION USING dEEP HYBRId REPRESENTATION LEARNING FOR 2d ANd 3d MICROSCOPY imageS 21

CELL NUCLEI SEGMENTATION USING DEEP HYBRID REPRESENTATION LE...

引用

21st IEEE International Symposium on Biomedical Imaging (ISBI)

作者： Meyer, J. dong, X. Rohr, K. Heidelberg Univ Biomed Comp Vision Grp BioQuant IPMB Heidelberg Germany

ISBN: (纸本)9798350313345;9798350313338

Cell segmentation is a central task in biomedical image analysis. We introduce a 3d deep neural network for 3d cell nuclei segmentation that performs multi-task learning to generate different representations from 3d microscopy images. The network exploits information about cell regions, cell contours, and cell distances. It consists of a 3d Residual U-Net with Squeeze-and-Excitation operations and is trained with a recently introduced loss function that suppresses image regions away from cell borders. To improve cell splitting, we propose a novel representation named signed Instance distance. Cell instance segmentation is achieved by combining the learned representations to determine cell seeds for a watershed algorithm. We also introduce a 2d version of the network. The method was applied to large-scale 3d micro-CT and 3d electron microscopy datasets from NucMM as well as to 2d fluorescence and phase-contrast microscopy datasets from the Cell Tracking Challenge. For the 3d datasets, the method achieves a large improvement compared to existing methods. For the 2d datasets, the method yields the best segmentation accuracy out of 39 methods for one of the datasets.

关键词： Biomedical imaging 3d image data cell instance segmentation deep learning electron microscopy micro-CT light microscopy

来源：评论

学校读者我要写书评

暂无评论

Cardiac Valve Annulus Manual Segmentation Using Computer Assisted Visual Feedback in Three-dimensional image data

Cardiac Valve Annulus Manual Segmentation Using Computer Ass...

引用

32nd Annual International Conference of the IEEE Engineering-in-Medicine-and-Biology-Society (EMBC 10)

作者： Casero, Ramon Burton, Rebecca A. B. Quinn, T. Alexander Bollensdorff, Christian Hales, Patrick Schneider, Juergen E. Kohl, Peter Grau, Vicente Univ Oxford Comp Lab Computat Biol Grp Wolfson BldgParks Rd Oxford OX1 3QD England Univ Oxford Cardiac Mechano Elect Feedback Lab Dept Physiol Anatomy & Genet Oxford OX1 3PT England Univ Oxford Wellcome Trust Ctr Human Genet Dept Cardiovascular Med Oxford OX3 7BN England Univ Oxford Inst Biomed Engn Dept Engn Sci Oxford OX1 3QD England

ISBN: (纸本)9781424441242

Annulus manual segmentation is an important tool for the study of valve anatomy and physiology, for the four main valves of the heart (mitral, tricuspid, aortic and pulmonary). In this paper we review two traditional manual segmentation approaches: slice-by-slice and interpolating a sparse set of landmarks with a spline curve. We propose a new Spline Tool for the open source software platform Seg3d, that is fast and improves spatial coherence by providing visual feedback of the segmentation in real time. The Spline Tool was tested successfully on 14 rat hearts, on all four valves.

关键词： 3d image data Seg3d open source software platform aortic valve cardiac valve anatomy cardiac valve annulus cardiac valve physiology cardiology computer assisted visual feedback image segmentation manual image segmentation medical image processing mitral valve pulmonary valve slice by slice manual segmentation sparse landmark set interpolation spline curve splines (mathematics) tricuspid valve

来源：评论

学校读者我要写书评

暂无评论

dEEP NEURAL NETWORK FOR 3d PARTICLE dETECTION IN 3d FLUORESCENCE MICROSCOPY imageS VIA dENSITY MAP REGRESSION 19

DEEP NEURAL NETWORK FOR 3D PARTICLE DETECTION IN 3D FLUORESC...

引用

19th IEEE International Symposium on Biomedical Imaging (IEEE ISBI)

作者： Spilger, R. Chagin, V. O. Bold, C. S. Schermelleh, L. Muller, U. C. Cardoso, M. C. Rohr, K. Heidelberg Univ Biomed Comp Vis Grp IPMB BioQuant Heidelberg Germany Tech Univ Darmstadt Cell Biol & Epigenet Dept Biol Darmstadt Germany Russian Acad Sci Inst Cytol St Petersburg Russia Heidelberg Univ IPMB Funct Genom Heidelberg Germany Univ Oxford Dept Biochem Micron Adv Bioimaging Unit Oxford England

ISBN: (纸本)9781665429238

Automatic detection of particles in fluorescence microscopy images is crucial to analyze cellular processes. We introduce a novel deep learning method for 3d fluorescent particle detection. Instead of pixel-wise binary classification or direct coordinate regression, we perform image-to-image mapping based on regressing a density map. detections close to particles are rewarded in the network training, and highly nonlinear direct prediction of point coordinates is avoided. To focus on particles in comparison to background image points, we suggest using the adaptive wing loss. We also employ a weighted loss map to cope with the very strong imbalance between particle and background image points for 3d images. We evaluated our approach using 3d images of the Particle Tracking Challenge and real 3d fluorescence microscopy images of chromatin structures and interneurons. It turned out that our approach generally outperforms previous methods.

关键词： Biomedical imaging fluorescence microscopy 3d image data particle detection deep learning

来源：评论

学校读者我要写书评

暂无评论

Rozšíření knihovny OpenSceneGraph pro zobrazení 3d obrazových dat v medicíně

Rozšíření knihovny OpenSceneGraph pro zobrazení 3D obra...

引用

作者： Kučiš, Michal Brno University of Technology

Bakalářská práce popisuje metody pro zobrazen37; 3d obrazov3;ch dat. Součast37; práce je implementace metody, která zobrazuje řez daty.

Bakalářská práce popisuje metody pro zobrazení 3d obrazových dat. Součastí práce je implementace metody, která zobrazuje řez daty.

关键词： OpenSceneGraph osg osgMedicalExt 3d obrazové data medicína řezová rovina OpenSceneGraph osg osgMedicalExt 3d image data medicine cut plane Text

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：