检索结果-内蒙古大学图书馆

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Pedro R. A. S. Bassi Wenxuan Li Yucheng Tang Fabian Isensee Zifu Wang Jieneng Chen Yu-Cheng Chou Saikat Roy Yannick Kirchhoff Maximilian Rokuss Ziyan Huang Jin Ye Junjun He Tassilo Wald Constantin Ulrich Michael Baumgartner Klaus H. Maier-Hein Paul Jaeger Yiwen Ye Yutong Xie Jianpeng Zhang Ziyang Chen Yong Xia Zhaohu Xing Lei Zhu Yousef Sadegheih Afshin Bozorgpour Pratibha Kumari Reza Azad Dorit Merhof Pengcheng Shi Ting Ma Yuxin Du Fan Bai Tiejun Huang Bo Zhao Haonan Wang Xiaomeng Li Hanxue Gu Haoyu Dong Jichen Yang Maciej A. Mazurowski Saumya Gupta Linshan Wu Jiaxin Zhuang Hao Chen Holger Roth Daguang Xu Matthew B. Blaschko Sergio Decherchi Andrea Cavalli Alan L. Yuille Zongwei Zhou Department of Computer Science Johns Hopkins University and Department of Pharmacy and Biotechnology University of Bologna and Center for Biomolecular Nanotechnologies Istituto Italiano di Tecnologia Department of Computer Science Johns Hopkins University NVIDIA Division of Medical Image Computing German Cancer Research Center (DKFZ) and Helmholtz Imaging German Cancer Research Center (DKFZ) ESAT-PSI KU Leuven Division of Medical Image Computing German Cancer Research Center (DKFZ) and Faculty of Mathematics and Computer Science Heidelberg University Division of Medical Image Computing German Cancer Research Center (DKFZ) and Faculty of Mathematics and Computer Science Heidelberg University and HIDSS4Health - Helmholtz Information and Data Science School for Health Shanghai Jiao Tong University Shanghai Artificial Intelligence Laboratory Division of Medical Image Computing German Cancer Research Center (DKFZ) Division of Medical Image Computing German Cancer Research Center (DKFZ) and Pattern Analysis and Learning Group Department of Radiation Oncology Heidelberg University Hospital Helmholtz Imaging German Cancer Research Center (DKFZ) and Interactive Machine Learning Group (IML) DKFZ School of Computer Science and Engineering Northwestern Polytechnical University Australian Institute for Machine Learning The University of Adelaide College of Computer Science and Technology Zhejiang University Hong Kong University of Science and Technology (Guangzhou) Hong Kong University of Science and Technology (Guangzhou) and Hong Kong University of Science and Technology Faculty of Informatics and Data Science University of Regensburg Faculty of Electrical Engineering and Information Technology RWTH Aachen University Faculty of Informatics and Data Science University of Regensburg and Fraunhofer Institute for Digital Medicine MEVIS Electronic & Information Engineering School Harbin Institute of Technology (Shenzhen) Shanghai Jiao Tong University and Beijing Academy of Artificial Intelligence (BAAI) S

ISBN: (纸本)9798331314385

How can we test AI performance? This question seems trivial, but it isn't. Standard benchmarks often have problems such as in-distribution and small-size test sets, oversimplified metrics, unfair comparisons, and short-term outcome pressure. As a consequence, good performance on standard benchmarks does not guarantee success in real-world scenarios. To address these problems, we present Touchstone, a large-scale collaborative segmentation benchmark of 9 types of abdominal organs. This benchmark is based on 5,195 training CT scans from 76 hospitals around the world and 5,903 testing CT scans from 11 additional hospitals. This diverse test set enhances the statistical significance of benchmark results and rigorously evaluates AI algorithms across out-of-distribution scenarios. We invited 14 inventors of 19 AI algorithms to train their algorithms, while our team, as a third party, independently evaluated these algorithms. In addition, we also evaluated pre-existing AI frameworks—which, differing from algorithms, are more flexible and can support different algorithms—including MONAI from NVIDIA, nnU-Net from DKFZ, and numerous other open-source frameworks. We are committed to expanding this benchmark to encourage more innovation of AI algorithms for the medical domain.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Automated detection of vertebral fractures in CT using 3D convolutional neural networks

引用

Bone Reports 2020年 13卷

作者： Joeri Nicolaes Steven Raeymaeckers David Robben Guido Wilms Dirk Vandermeulen Cesar Libanati Marc Debois UCB Pharma Brussels Belgium Medical Image Computing ESAT-PSI Department of Electrical Engineering KU Leuven Leuven Belgium Department of Radiology University Hospital Brussels Brussels Belgium icometrix Leuven Belgium Department of Radiology KU Leuven Leuven Belgium

来源：评论

学校读者我要写书评

暂无评论

Groupwise deformable registration of fiber track sets using track orientation distributions

Groupwise deformable registration of fiber track sets using ...

引用

16th International Conference on medical image computing and Computer Assisted Intervention, MICCAI 2013

作者： Christiaens, Daan Dhollander, Thijs Maes, Frederik Sunaert, Stefan Suetens, Paul Department of Electrical Engineering ESAT/PSI Medical Image Computing Medical Imaging Research Center KU Leuven Herestraat 49 Leuven Belgium Department of Imaging & Pathology Translational MRI Medical Imaging Research Center KU Leuven Herestraat 49 Leuven Belgium

Diffusion-weighted imaging (DWI) and tractography allow to study the macroscopic structure of white matter in vivo. We present a novel method for deformable registration of unsegmented full-brain fiber track sets extracted from DWI data. Our method attempts to align the track orientation distributions (TODs) of multiple subjects, rather than individual tracks. As such, it can handle complex track configurations and allows for multi-resolution registration. We validated the registration method on synthetically deformed DWI data and on data of 15 healthy subjects, and achieved sub-voxel accuracy in dense white matter structures. This work is, to the best of our knowledge, the first demonstration of direct registration of probabilistic tractography data. © Springer International Publishing Switzerland 2014.

关键词： Deformation

来源：评论

学校读者我要写书评

暂无评论

Validating cone-beam computed tomography for peri-implant bone morphometric analysis

引用

Bone Research 2014年第2期2卷 103-109页

作者： Yan Huang Jeroen Van Dessel Maarten Depypere Mostafa EzEldeen Alexandru Andrei Iliescu Emanuela Dos Santos Ivo Lambrichts Xin Liang Reinhilde Jacobs OMFS IMPATH Research Group Department of Imaging & Pathology Faculty of Medicine University of Leuven and Oral & Maxillofacial SurgeryUniversity Hospitals Leuven Leuven Belgium State Key Laboratory of Oral Diseases West China Hospital of Stomatology Sichuan University ESAT/PSI Medical Image Computing KU Leuven Leuven Belgium Faculty of Dentistry UMF Craiova Romania School of Dentistry Pontifical Catholic University of Paraná Curitiba Brazil Biomedical Research Institute Hasselt University Diepenbeek Belgium School of Stomatology Dalian Medical University Dalian China

Cone-beam computed tomography （CBCT） has been recently used to analyse trabecular bone structure around dental implants. To validate the use of CBCT for three-dimensional （3D） peri-implant trabecular bone morphometry by comparing it to two-dimensional （2D） histology, 36 alveolar bone samples （with implants n=27 vs. without implants n=9） from six mongrel dogs, were scanned ex vivo using a high-resolution （80 ~m） CBCT. After scanning, all samples were decalcified and then sectioned into thin histological sections （-6 jim） to obtain high contrast 2D images. By using CTAn imaging software, bone morphometric parameters including trabecular number （Tb.N）, thickness （***）, separation （***） and bone volume fraction （BV/TV） were examined on both CBCT and corresponding histological images. Higher *** and ***, lower BV/TV and Tb.N were found on CBCT images （P〈0.001）. Both measurements on the peri-implant trabecular bone structure showed moderate to high correlation （r=0.65-0.85）. The Bland-Altman plots showed strongest agreement for *** followed by ***, Tb.N and BV/TV, regardless of the presence of implants. The current findings support the assumption that peri-implant trabecular bone structures based on high-resolution CBCT measurements are representative for the underlying histological bone characteristics, indicating a potential clinical diagnostic use of CBCT-based peri-implant bone morphometric characterisation.

关键词： bone CBCT

来源：评论

学校读者我要写书评

暂无评论

An automated pipeline for regional cardiac strain estimation from volumetric ultrasound data

An automated pipeline for regional cardiac strain estimation...

引用

IEEE International Ultrasonics Symposium

作者： Brecht Heyde Daniel Barbosa Ana-Maria Daraban Ruta Jasaityte Piet Claus Frederik Maes Jan D'hooge Cardiovascular Imaging and Dynamics University of Leuven ESAT/PSI - Medical Image Computing University of Leuven

ISBN: (纸本)9781467356855

Strain estimation techniques can provide valuable insights into regional cardiac (dys)function. However, a manual delineation of the left ventricle is required, intrinsically leading to a time consuming analysis with significant observer variability. In this study we therefore propose a fully automated strain estimation strategy with minimal user input, and illustrate its applicability in a realistic clinical setting. To test the proposed approach, volumetric data (32±3 Hz) was recorded in 3 healthy volunteers and 3 patients with acute myocardial infarction. User interaction time was considerably shorter than manual contouring (19±3s vs 390±90s) and eliminated observer variability. The fully automated strain estimation pipeline showed strain estimates in both subject groups that were in agreement with conventional analysis. Comparing the infarct regions against the remote regions in the patient group, end-systolic strain was reduced by 13.7%, 3.8% and 8.5% for the automated method, and by 12.3%, 5.0% and 9.0% if manually contoured, for the radial, longitudinal and circumferential strain component respectively. Furthermore, the strain curves of both investigated strategies matched well. The proposed automated approach is an attractive solution for clinical practice.

关键词： Strain Echocardiography Automated Elastic registration Segmentation

来源：评论

学校读者我要写书评

暂无评论

1.2. A novel approach to automated bloodstain pattern analysis using an active bloodstain shape model

引用

Journal of Forensic Radiology and Imaging 2014年第2期2卷 95-95页

作者： P. Joris W. Develter E. Jenar D. Vandermeulen W. Coudyzer J. Wuestenbergs B. De Dobbelaer W. Van de Voorde E. Geusens P. Claes University of Leuven (KUL) Department of Electrical Engineering ESAT/PSI Medical Image Computing & UZ Leuven Medical Imaging Research Center & iMinds Future Health Department Belgium University Hospitals Leuven (UZ) Department of Forensic Medicine Belgium University Hospitals Leuven (UZ) Department of Radiology Belgium

Objectives Conventional blood pattern analysis is a tedious and time- consuming process due to the many actions that must be performed by the pattern analyst. Digital photographs can now be employed to aid in the process of scene analysis, by letting computer programs perform some of the required steps. However, current computerized methods are all based upon the assumption that any stain can be approximated by a simple ellipse, while manual work is still required. Materials and methods This work presents a novel approach, employing a regressed active shape model to approximate bloodstains. This Active Bloodstain Shape Model (ABSM) uses a regression to correlate the shape of a stain׳s approximation to its impact angle. This model is then deployed in a software pipeline, aiming to eliminate any user-input from the process. Fiducial markers are positioned in a crime scene, allowing for fully automated pattern analysis. images are cleaned of any perspective distortions, after which stains are segmented and analyzed using said model. A robust principal component analysis (PCA) is then used to analyze the intersections of the reconstructed flight paths. Results Experimental results have demonstrated that the ABSM is able to better approximate stains, while accuracy on predicted impact angles is increased. Both simulated as well as real crimes scenes have confirmed the working of the automated pipeline. Conclusion Because the ABSM learns from experimental data, it is able to better approximate stains, resulting in an increased accuracy. The addition of the software pipeline removed the need for almost all user-input.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3.5. The feasibility of measuring body weight on CT images and the first steps in anatomical mirroring

引用

Journal of Forensic Radiology and Imaging 2014年第2期2卷 105-105页

作者： J. Wuestenbergs W. Coudyzer B. De Dobbelaer W. Develter P. Joris P. Claes D. Vandermeulen E. Jenar E. Geusens W. Van de Voorde University Hospitals Leuven (UZ) Department of Forensic Medicine Belgium University Hospitals Leuven (UZ) Department of Radiology Belgium University of Leuven (KUL) Belgium Department of Electrical Engineering ESAT/PSI Medical Image Computing & UZ Leuven Medical Imaging Research Center & iMinds Future Health Department Belgium

Objective To estimate body weight using the data of a post mortem total body CT scan and establish its efficacy in a forensic context. Material and method PMCT data were used to compute fat, soft tissue and bone volume using in house developed software based on density and HU. The digitally calculated body weight was then compared with the measured body weight on autopsy. Results Analysis of the obtained results showed a good correlation between the measured body weight and the CT-derived body weight. The highest discrepancies were noticed in children and decomposed bodies presumably due to different bone mineralization in children versus adults and putrefactive gas formation and liquefaction in decomposed bodies. In general, there was no distinct consistent over- or underestimation. In case of partial scans, where a body part was not scanned or was missing, mirroring the unaffected side was seen to give equally good results. Conclusion The weight of a body can be sufficiently estimated using volume measurements of different body tissues obtained from the PMCT data. This technique can be applied when concern is raised about the documented body weight, or if weighing was omitted all together. Furthermore the aforementioned body weight estimates can be useful in case of mass casualties, for victim identification purposes, or in case of traumatic deaths (e.g. dismemberments).

关键词：

来源：评论

学校读者我要写书评

暂无评论

Three-dimensional myocardial strain estimation from volumetric ultrasound data using a novel transformation model adapted to the heart

Three-dimensional myocardial strain estimation from volumetr...

引用

IEEE International Ultrasonics Symposium

作者： Brecht Heyde Daniel Barbosa Ruta Jasaityte Stefaan Bouchez Patrick Wouters Frederik Maes Piet Claus Jan D'hooge Cardiovascular Imaging and Dynamics University of Leuven Leuven Belgium Department of Anesthesiology Ghent University Ghent Belgium ESAT/PSI - Medical Image Computing University of Leuven Leuven Belgium

ISBN: (纸本)9781467345613

We previously demonstrated that 3D cardiac strain estimation using non-rigid registration based on a B-spline freeform deformation model (FFD) is feasible. However, the traditional arrangement of the B-spline control points on a Cartesian grid (CFFD) may be suboptimal for cardiac applications as it treats the blood pool and myocardium similarly and as it enforces smoothness in non-physiologic directions. The aim of this study was therefore to overcome these limitations by proposing a novel anatomical FFD (AFFD) adapted to the heart and test its performance experimentally. Volumetric data was recorded in gel phantoms and in 5 sheep. Reference radial (∈_(RR)), longitudinal (∈_(LL)) and circumferential strain (∈_(CC)) were obtained using sonomicrometry. The strain range was modulated by increasing the stroke volume (phantoms;25-150ml) or by (pharmacological/surgical) inotropic modulation (sheep). Correlation coefficients for ∈_(RR), ∈_(LL) and ∈_(CC) were 0.98, 0.62 and 0.94 respectively in the phantom data and 0.87, 0.65 and 0.74 respectively in the sheep data. Moreover, the shape of the strain curves, timing of peak values and location of dysfunctional regions were recovered well. Further validation and its comparison to the CFFD model is the topic of ongoing research.

关键词： Non-rigid registration anatomical deformation model sonomicrometry strain estimation echocardiography Heart Echocardiography Imaging Phantom nonrigidity Myocardium splines (mathematics) Paranormal phenomena control point Ovis Smoothness

来源：评论

学校读者我要写书评

暂无评论

CBCT instead of microCT for studying the trabecular bone:a feasible option?

CBCT instead of microCT for studying the trabecular bone:a f...

引用

The 9th Asian Congress of Oral and Maxillofacial Radiology

作者： Jeroen Van Dessel Maarten Depypere Bart Vandenberghe Frederik Maes Ivo Lambrichts Reinhilde Jacobs Oral Imaging Center Department of Oral Health SciencesKU LeuvenLeuvenBelgium ESAT/PSI Medical Image ComputingKULeuvenLeuvenBelgium Department of Morphology Faculty of MedicineUniversity of HasseltDiepenbeekBelgium

Aims:The feasibility of using CBCT images in analysing trabecular bone structure parameters at peri-implant regions were discussed,and to compare trabecular bone parameters obtaind by CBCT and micro-CT.

关键词： CBCT instead of microCT for studying the trabecular bone:a feasible option

来源：评论

学校读者我要写书评

暂无评论

Impact of RF inhomogeneity correction on image registration of micro MRI rodent brain images

Impact of RF inhomogeneity correction on image registration ...

引用

IEEE International Symposium on Biomedical Imaging

作者： J.R. Rangarajan D. Loeckx G. Vande Velde T. Dresselaers U. Himmelreich F. Maes P. Suetens Medical Image Computing-ESAT PSI Biomedical NMR Unit Katholieke Universiteit Leuven Belgium

High field strength small animal magnetic resonance imaging (microMRI) allows to visualize the internal anatomy of the animal, especially of the brain, with high resolution. Longitudinal in vivo small animal molecular imaging applications greatly benefit from advanced image processing methods for removing image artifacts like inhomogeneity, to spatially normalize images and to obtain atlas based segmentation. Subtle interactions between different methods, which might otherwise remain undiscovered if the individual methods are validated with phantom or simulated data, are better evaluated using real-world data. In this paper, we quantitatively demonstrate the confounding effect of RF inhomogeneity on image registration of in vivo microMRI images of rodent brain and validate the benefit of bias field correction on both registration accuracy and registration robustness.

关键词： Animals image segmentation Brain Nonhomogeneous media In vivo Pipelines Imaging

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：