专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

WO2008097520A3 SYSTEM AND METHOD FOR COMPUTER AIDED DETECTION OF PULMONARY EMBOLISM IN TOBOGGANING IN CT ANGIOGRAPHY 审中-公开
标题翻译：计算机辅助检测CT血管造影中肺动脉栓塞的系统与方法
公开(公告)号：WO2008097520A3
公开(公告)日：2009-06-04
申请号：PCT/US2008001474
申请日：2008-02-04
申请人： SIEMENS MEDICAL SOLUTIONS , LIANG JIANMING , BI JINBO
发明人： LIANG JIANMING , BI JINBO
IPC分类号： G06T7/00 , A61B6/03 , G06K9/00
CPC分类号： G06K9/00201 , G06T7/0012 , G06T7/11 , G06T2207/10081 , G06T2207/30061 , G06T2207/30101
摘要： A method for detecting pulmonary embolisms in computed tomographic (CT) angiography includes providing a digitized (CT) image acquired from inside a pulmonary vessel, the image comprising a plurality of intensities corresponding to a 3-dimensional grid of voxels, for each voxel in the image, extracting a first pulmonary embolism (PE) candidate and PE boundary from the voxel, and for each voxel in the PE boundary, selecting a voxel from the PE boundary, extracting a subsequent PE candidate and PE boundary from the voxel, merging the subsequent PE candidate with the first PE candidate, and merging the subsequent PE boundary with the first PE boundary.
摘要翻译：用于检测计算机断层摄影（CT）血管造影术中的肺栓塞的方法包括提供从肺容器内获得的数字化（CT）图像，该图像包括对应于体素的三维网格的多个强度，图像，从体素提取第一肺栓塞（PE）候选者和PE边界，并且对于PE边界中的每个体素，从PE边界中选择体素，从体素提取后续PE候选和PE边界，合并后续 PE候选人与第一个PE候选人，并将后续PE边界与第一个PE边界合并。

2. 发明申请

WO2007033228A2 REDUCING FALSE POSITIVES FOR AUTOMATIC COMPUTERIZED DETECTION OF OBJECTS 审中-公开
标题翻译：减少对物体自动计算机检测的虚假信息
公开(公告)号：WO2007033228A2
公开(公告)日：2007-03-22
申请号：PCT/US2006/035601
申请日：2006-09-11
申请人： SIEMENS MEDICAL SOLUTIONS USA, INC. , LIANG, Jianming , XUE, Zhiyun , SALGANICOFF, Marcos , WOLF, Matthias
发明人： LIANG, Jianming , XUE, Zhiyun , SALGANICOFF, Marcos , WOLF, Matthias
CPC分类号： G06T7/0012
摘要： A computer-implemented method for identifying an object of interest includes providing input data including an image and a candidate for the object of interest in the image (201), extracting a boundary of the candidate, and extracting a segment of a region of interest containing the candidate (202). The method further includes determining a plurality of features of an extracted segment of the region of interest containing the candidate (203), and outputting the object of interest (204), wherein the object of interest is characterized by the plurality of features (205), wherein the object of interest and the plurality of features are stored as computer-readable code.
摘要翻译：用于识别感兴趣对象的计算机实现的方法包括：提供包括图像的图像和候选对象的输入数据（201），提取候选的边界，以及提取包含候选人（202）。该方法还包括确定包含候选（203）的感兴趣区域的提取段的多个特征，并输出感兴趣对象（204），其中感兴趣对象由多个特征（205）表征，，其中感兴趣对象和多个特征被存储为计算机可读代码。

3. 发明申请

WO2005062255A1 TOBOGGAN-BASED SHAPE CHARACTERIZATION 审中-公开
标题翻译：基于TOBOGGAN的形状特征
公开(公告)号：WO2005062255A1
公开(公告)日：2005-07-07
申请号：PCT/US2004/041116
申请日：2004-12-08
申请人： SIEMENS MEDICAL SOLUTIONS USA, INC. , LIANG, Jianming , BOGONI, Luca
发明人： LIANG, Jianming , BOGONI, Luca
IPC分类号： G06T5/00
CPC分类号： G06T7/0012 , G06K9/342 , G06K2209/05 , G06T7/11 , G06T7/155 , G06T2207/10068 , G06T2207/20152 , G06T2207/30032
摘要： A method (900) and apparatus (101) for characterizing an image. The method selects one or more toboggan potentials from the image, or a portion thereof, to be tobogganed (920). It toboggans the selected toboggan potentials to generate one or more toboggan parameters (930), forming at least one toboggan cluster using one or more of the toboggan parameters (940). It also selects one or more of the toboggan clusters to compute at least one feature parameter to characterize the image or a portion thereof (950).
摘要翻译：一种用于表征图像的方法（900）和装置（101）。该方法从图像或其一部分中选择一个或多个滑降电位来进行滑行（920）。它使所选择的滑降电位滚动以生成一个或多个滑降参数（930），使用一个或多个滑降参数形成至少一个滑降簇（940）。它还选择一个或多个滑降簇来计算至少一个特征参数以表征图像或其一部分（950）。

4. 发明申请

WO2008097520A2 SYSTEM AND METHOD FOR COMPUTER AIDED DETECTION OF PULMONARY EMBOLISM IN TOBOGGANING IN CT ANGIOGRAPHY 审中-公开
标题翻译：计算机辅助检测CT血管造影中肺动脉栓塞的系统与方法
公开(公告)号：WO2008097520A2
公开(公告)日：2008-08-14
申请号：PCT/US2008/001474
申请日：2008-02-04
申请人： SIEMENS MEDICAL SOLUTIONS USA, INC. , LIANG, Jianming , BI, Jinbo
发明人： LIANG, Jianming , BI, Jinbo
IPC分类号： G06T7/00 , A61B6/03 , A61B6/12
CPC分类号： G06K9/00201 , G06T7/0012 , G06T7/11 , G06T2207/10081 , G06T2207/30061 , G06T2207/30101
摘要： A method for detecting pulmonary embolisms in computed tomographic (CT) angiography includes providing a digitized (CT) image acquired from inside a pulmonary vessel, the image comprising a plurality of intensities corresponding to a 3-dimensional grid of voxels, for each voxel in the image, extracting a first pulmonary embolism (PE) candidate and PE boundary from the voxel, and for each voxel in the PE boundary, selecting a voxel from the PE boundary, extracting a subsequent PE candidate and PE boundary from the voxel, merging the subsequent PE candidate with the first PE candidate, and merging the subsequent PE boundary with the first PE boundary.
摘要翻译：用于检测计算机断层摄影（CT）血管造影术中的肺栓塞的方法包括提供从肺容器内获得的数字化（CT）图像，该图像包括对应于体素的三维网格的多个强度，图像，从体素提取第一肺栓塞（PE）候选者和PE边界，并且对于PE边界中的每个体素，从PE边界中选择体素，从体素提取后续PE候选和PE边界，合并后续 PE候选人与第一个PE候选人，并将后续PE边界与第一个PE边界合并。

5. 发明申请

WO2006042322A1 A TOBOGGAN-BASED METHOD FOR AUTOMATIC DETECTION AND SEGMENTATION OF OBJECTS IN IMAGE DATA 审中-公开
标题翻译：一种基于TOBOGGAN的方法，用于自动检测和分割图像数据中的对象
公开(公告)号：WO2006042322A1
公开(公告)日：2006-04-20
申请号：PCT/US2005/037051
申请日：2005-10-12
申请人： SIEMENS MEDICAL SOLUTIONS USA, INC. , LIANG, Jianming , WOLF, Matthias , SALGANICOFF, Marcos
发明人： LIANG, Jianming , WOLF, Matthias , SALGANICOFF, Marcos
IPC分类号： G06T5/30
CPC分类号： G06K9/342 , G06K2209/05 , G06T7/11 , G06T7/155 , G06T2207/10081 , G06T2207/20152 , G06T2207/30032 , G06T2207/30064 , G06T2207/30101
摘要： An exemplary method of detecting one or more objects in image data is provided. The image data includes a plurality of pixels/voxels. The method includes sliding (at 810) pixels/voxels that meet sliding criteria; and collecting (at 815) the slid pixels/voxels that satisfy collecting criteria. An exemplary method of segmenting an object in image data is also provided. The method includes receiving an initial pixel/voxel in the image data; and forming a segmentation of the object based on the initial pixel/voxel.
摘要翻译：提供了检测图像数据中的一个或多个对象的示例性方法。图像数据包括多个像素/体素。该方法包括滑动（在810处）符合滑动标准的像素/体素; 并收集（815）满足收集标准的滑动像素/体素。还提供了分割图像数据中的对象的示例性方法。该方法包括在图像数据中接收初始像素/体元; 并基于初始像素/体元形成对象的分割。

6. 发明申请

WO2006041549A1 MEDICAL DIAGNOSTIC ULTRASOUND CHARACTERIZATION OF CARDIAC MOTION 审中-公开
标题翻译：心脏运动的医学诊断超声波特征
公开(公告)号：WO2006041549A1
公开(公告)日：2006-04-20
申请号：PCT/US2005/026442
申请日：2005-07-26
申请人： SIEMENS MEDICAL SOLUTIONS USA, INC. , LIANG, Jianming , KRISHNAN, Sriram , RAO, R. Bharat
发明人： LIANG, Jianming , KRISHNAN, Sriram , RAO, R. Bharat
IPC分类号： A61B8/08 , A61B8/02 , G01S15/89 , G06T7/00
CPC分类号： G01S15/8981 , A61B8/02 , A61B8/08 , A61B8/0866 , A61B8/0883 , A61B2018/00351 , G06K9/6207 , G06K2209/05 , G06T7/0012 , G06T7/20 , G06T2207/30048
摘要： Motion is automatically characterized (34) from ultrasound information. Ultrasound information associated with particular time periods relative to a cycle, such as the cardiac cycle, are extracted, such as identifying and extracting ultrasound information associated with systole (26, 28). By tracking (22) an area (20) of the heart, such as an area within the endocardial contour, the heart cycle time periods are identified (24). Spatial parameter values are determined (32) as a function of time from the extracted ultrasound information. For example, the timing of motion, the eigen motion, the curvature, the local ejection-fraction ratio and/or the bending energy of parts of the cardiac tissue are determined (32). The spatial parameter values characterize (34) the motion.
摘要翻译：运动根据超声信息自动表征（34）。提取与相对于循环（例如心动周期）的特定时间段相关联的超声信息，例如识别和提取与收缩相关的超声信息（26,28）。通过跟踪（22）心脏区域（20）（例如心内膜轮廓内的区域），识别心脏周期时间段（24）。从提取的超声信息中确定空间参数值（32）作为时间的函数。例如，确定心脏组织部分的运动时间，特征运动，曲率，局部射血分数比和/或弯曲能量（32）。空间参数值表征（34）运动。

7. 发明申请

WO2006019547A1 SYSTEM AND METHOD FOR OBJECT CHARACTERIZATION OF TOBOGGAN-BASED CLUSTERS 审中-公开
标题翻译：基于TOBOGGAN的簇的对象特征的系统和方法
公开(公告)号：WO2006019547A1
公开(公告)日：2006-02-23
申请号：PCT/US2005/023558
申请日：2005-07-05
申请人： SIEMENS MEDICAL SOLUTIONS USA, INC. , LIANG, Jianming , BOGONI, Luca
发明人： LIANG, Jianming , BOGONI, Luca
IPC分类号： G06T5/00
CPC分类号： G06T7/0012 , G06K9/00 , G06K9/342 , G06K9/46 , G06K2209/05 , G06T7/11 , G06T7/155 , G06T2207/10072 , G06T2207/20152 , G06T2207/30004
摘要： A method for characterizing an object in a digitized image includes providing (41) a digitized volumetric image comprising a plurality of intensities corresponding to a domain of points in an N-dimensional space, forming (42) a toboggan cluster from a subset of contiguous points in said image, said toboggan cluster including a concentration point, extracting (44) a first layer from said toboggan cluster, and computing (45) one or more features from said toboggan layer
摘要翻译：用于表征数字化图像中的对象的方法包括提供（41）数字化体积图像，其包括对应于N维空间中的点的域的多个强度，从连续点的子集中形成（42）滑降簇在所述图像中，所述滑冰簇包括集中点，从所述滑冰簇提取（44）第一层，以及计算（45）来自所述滑冰层的一个或多个特征

8. 发明申请

WO2005122088A1 SYSTEM AND METHOD FOR DYNAMIC FAST TOBOGGANING 审中-公开
标题翻译：动态快速TOBOGGANING的系统与方法
公开(公告)号：WO2005122088A1
公开(公告)日：2005-12-22
申请号：PCT/US2005/020069
申请日：2005-06-07
申请人： SIEMENS MEDICAL SOLUTIONS USA, INC. , LIANG, Jianming , BOGONI, Luca
发明人： LIANG, Jianming , BOGONI, Luca
IPC分类号： G06T5/00
CPC分类号： G06K9/342 , G06T7/11 , G06T7/155 , G06T2207/10081 , G06T2207/20152 , G06T2207/20164 , G06T2207/30032
摘要： A method of identifying an object in a digital image includes finding (41) a point in a digital image that is a concentration location, initializing (42) a cluster with said concentration location, adding the neighboring points of the concentration location to a list, selecting (43) a neighbor point with an extremal potential value from said list, determining a slide direction of all neighbors of said selected point and identifying (44) those neighbors that slide to the selected point, adding (45) those neighbor points not already in the list to the list, adding the selected point to the cluster, and repeating(46) the steps of selecting a neighbor point with an extremal potential value, determining a slide direction, adding points to the list, and adding the selected point to the cluster, until the list is empty
摘要翻译：一种识别数字图像中的对象的方法包括：找到（41）作为浓度位置的数字图像中的点，初始化（42）具有所述浓度位置的聚类，将浓度位置的相邻点添加到列表中，从所述列表中选择（43）具有极值电位值的邻居点，确定所述所选点的所有邻居的滑动方向，并识别（44）滑动到所选点的那些邻居，添加（45）尚未存在的相邻点在列表中将列表添加到集群中，并重复（46）选择具有极值电位值的邻居点的步骤，确定滑动方向，向列表添加点，并将所选择的点添加到集群，直到列表为空

9. 发明申请

WO2005122087A1 SYSTEM AND METHOD FOR TOBOGGAN-BASED OBJECT SEGMENTATION USING DISTANCE TRANSFORM 审中-公开
标题翻译：使用基于距离变换的基于TOBOGGAN的对象分割的系统和方法
公开(公告)号：WO2005122087A1
公开(公告)日：2005-12-22
申请号：PCT/US2005/020068
申请日：2005-06-07
申请人： SIEMENS MEDICAL SOLUTIONS USA, INC. , LIANG, Jianming , BOGONI, Luca
发明人： LIANG, Jianming , BOGONI, Luca
IPC分类号： G06T5/00
CPC分类号： G06T7/0012 , G06K9/342 , G06T7/11 , G06T7/155 , G06T2207/10081 , G06T2207/20041 , G06T2207/20152 , G06T2207/30028
摘要： A method of segmenting an object in a digital image comprising providing a digital image comprising a plurality of intensities corresponding to a domain of points in a N-dimensional space, selecting (91) a region of interest in the image, determining (92) a threshold intensity value for points in said region of interest, wherein an object of interest is defined by points with an intensity above a first pre-determined threshold, computing (93) a distance map for each point in said object of interest, tobogganing (94) each point in said object of interest based on said distance map, and selecting (95) a cluster based on the results of said tobogganing
摘要翻译：一种分割数字图像中的对象的方法，包括提供包括对应于N维空间中的点的域的多个强度的数字图像，选择（91）图像中的感兴趣区域，确定（92）所述感兴趣区域中的点的阈值强度值，其中感兴趣对象由强度高于第一预定阈值的点定义，计算（93）所述感兴趣对象中的每个点的距离图，雪橇（94 ）基于所述距离图的所述感兴趣对象中的每个点，并且基于所述滑冰的结果来选择（95）群集

10. 发明申请

WO2021016087A1 SYSTEMS FOR THE GENERATION OF SOURCE MODELS FOR TRANSFER LEARNING TO APPLICATION SPECIFIC MODELS 审中-公开
公开(公告)号：WO2021016087A1
公开(公告)日：2021-01-28
申请号：PCT/US2020/042560
申请日：2020-07-17
申请人： ARIZONA BOARD OF REGENTS ON BEHALF OF ARIZONA STATE UNIVERSITY , ZHOU, Zongwei , SODHA, Vatsal , RAHMAN SIDDIQUEE, Md, Mahfuzur , FENG, Ruibin , TAJBAKHSH, Nima , LIANG, Jianming
发明人： ZHOU, Zongwei , SODHA, Vatsal , RAHMAN SIDDIQUEE, Md, Mahfuzur , FENG, Ruibin , TAJBAKHSH, Nima , LIANG, Jianming
IPC分类号： G06F3/0481
摘要： Described herein are means for generating source models for transfer learning to application specific models used in the processing of medical imaging. In some embodiments, the method comprises: identifying training samples, wherein each training sample in the training samples includes an image; for each training sample: identifying an original patch of the image corresponding to the training sample; identifying one or more transformations to be applied to the original patch; generating a transformed patch by applying the one or more transformations to the identified patch; and training an encoderdecoder network using transformed patches corresponding to the training samples, wherein the encoder-decoder network is trained to generate an approximation of the original patch from a corresponding transformed patch, and wherein the encoder-decoder network is trained to minimize a loss function that indicates a difference between the generated approximation of the original patch and the original patch. The source models enhance the transfer learning performance for medical imaging tasks.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式