-
公开(公告)号：KR101479845B1
公开(公告)日：2015-01-09
申请号：KR1020130062311
申请日：2013-05-31
Applicant: 서울여자대학교 산학협력단
Abstract: 본 발명은 형상 사전정보 기반의 그래프 컷을 이용하여, 유사한 밝기 값의 주변 연부조직으로의 누출을 막고 전방 십자인대의 분할 성능을 크게 향상시킬 수 있는 무릎 자기공명영상의 전방십자인대 분할 기술에 관한 것으로, 본 발명에 따른 무릎 자기공명영상의 전방십자인대 분할 방법은 (a) 씨앗정보 추출부를 이용하여, 전체 영상 밝기 분포에서 가우시안 혼합 모델 근사를 수행하여, 어두운 밝기값 영역의 평균치 및 밝은 밝기값 영역의 평균치를 임계치로 설정하여, 씨앗 정보의 객체 후보군 및 배경 후보군을 추출하는 단계, (b) 씨앗정보 추출부를 이용하여, 상기 객체 후보군 및 배경 후보군에 형태학적 연산을 이용하여, 최종 객체 씨앗 정보 및 배경 씨앗 정보를 추출하는 단계 및 (c) 그래프컷 분할부를 이용하여, 추출된 상기 객체 씨앗 정보 및 배경 씨앗 정보로 형상 정보 기반의 그래프 컷을 분할하여, 최종 전방십자인대 볼륨을 분할하는 단계를 포함하는 것을 기술적 특징으로 한다.
-
公开(公告)号：KR1020140141055A
公开(公告)日：2014-12-10
申请号：KR1020130062311
申请日：2013-05-31
Applicant: 서울여자대학교 산학협력단
Abstract: The present invention relates to a technology for segmenting an anterior cruciate ligament from a knee MR image, which can prevent light having similar brightness values from leaking to surrounding soft tissues and significantly improve segmentation performance of the anterior cruciate ligament by using a shape advance information-based graph cut. According to the present invention, the method for segmenting the anterior cruciate ligament from the knee MR image comprises steps of: (a) performing gaussian mixture model estimation in the whole image brightness distribution to set a mean value of a dark brightness value region and a mean value of a bright brightness value region as critical values and extract an object candidate and a background candidate of seed information; (b) extracting final object seed information and background seed information through a morphological operation on the object candidate and the background candidate by using seed information extraction unit; and (c) segmenting the shape advance information-based graph cut into the extracted object seed information and the background seed information by using a graph cut segmentation unit to segment a final anterior cruciate ligament volume.
Abstract translation: 本发明涉及一种用于从膝关节MR图像分割前交叉韧带的技术，其可以防止具有相似亮度值的光泄露到周围的软组织，并且通过使用形状提前信息 - 显示来改善前交叉韧带的分割性能， 基于图形切割。 根据本发明，从膝盖MR图像分割前十字韧带的方法包括以下步骤：（a）在整个图像亮度分布中执行高斯混合模型估计，以设置暗亮度值区域和 将明亮度值区域的平均值作为临界值，并提取候选对象和种子信息的背景候选; （b）通过使用种子信息提取单元，通过对对象候选者和背景候选人的形态学操作来提取最终对象种子信息和背景种子信息; 以及（c）通过使用图形切割分割单元来分割切割成所提取的对象种子信息和背景种子信息的形状提前信息图，以分割最终的前十字韧带体积。
-
公开(公告)号：KR1020140054690A
公开(公告)日：2014-05-09
申请号：KR1020120120483
申请日：2012-10-29
Applicant: 서울여자대학교 산학협력단
CPC classification number: G06T7/33 , G06T5/002 , G06T5/50 , G06T2207/10088 , G06T2207/30096
Abstract: The present invention relates to a fusion system of T2-emphasized MR image and diffusion emphasized MR image and a method thereof capable of automatically detecting the location of malignant tumor by correcting rotation conversion distortion, location and size through rigid body matching of T2-emphasized MR image and diffusion emphasized MR image. The fusion method of T2-emphasized MR image and diffusion emphasized MR image comprises the following steps: strengthening similarity between brightness value signal distribution of two images and removing noise to improve the accuracy of conformation of two images of which the brightness value signal distribution and image resolution are different from each other; repeatedly performing rigid body matching for maximizing the normalization mutual information to correct the rotation conversion distortion, location and size of two images of which the similarity is strengthened; producing diffusion coefficient map into a color map to easily recognize malignant tumor in the adjusted image, and mapping the same to a tumor candidate group earned from the T2-emphasized MR image to coordinate the same to the T2-emphasized MR image.
Abstract translation: 本发明涉及T2强调MR图像和扩散强调MR图像的融合系统及其方法，其能够通过T2强调MR的刚体匹配来校正旋转转换失真，位置和大小来自动检测恶性肿瘤的位置 图像和扩散强调MR图像。 T2强调MR图像和扩散强调MR图像的融合方法包括以下步骤：加强两个图像的亮度值信号分布之间的相似性，并消除噪声，提高亮度值信号分布和图像的两个图像的构象精度 分辨率彼此不同; 重复执行刚体匹配，以最大化归一化互信息，以校正相似性加强的两个图像的旋转转换失真，位置和大小; 将扩散系数图生成到颜色图中以容易地识别调整图像中的恶性肿瘤，并将其映射到从T2强调的MR图像获得的肿瘤候选组，以将其与T2强调的MR图像协调。
-
44.发明授权동적조영증강 유방 ＭＲ 영상에서 조직별 밝기값 보정 및 종양 강성 제약을 적용한 비강체 정합 방법 및 시스템 有权非激素注射方法和系统，每个组织的密度校正和动态对比增强乳腺图像中肿瘤的刚度约束
公开(公告)号：KR101258814B1
公开(公告)日：2013-04-26
申请号：KR1020110116563
申请日：2011-11-09
Applicant: 서울여자대학교 산학협력단
Abstract: PURPOSE: A nonrigid registration method and system applying the tumor rigidity constraint and the density correction of each tissue in a dynamic contract enhanced breast MR image are provided to accurately classify a tissue by regularizing a contrast density. CONSTITUTION: A breast region is automatically divided by removing a breast skin and a pectoral(S110). A fuzzy C-means grouping method is used in the divided breast region. A breast tissue is classified into fat, the mammary gland, a tumor, and a blood vessel(S120). The tumor is divided by using a density and a morphological operation(S130). The density is corrected in the fat and the mammary gland(S140). [Reference numerals] (AA) Start; (BB) End; (S110) Breast division based on slope and direction information; (S120) Tissue classification using a fuzzy C-means grouping method; (S130) Tumor division based on brightness value and shape information; (S140) Rigid body matching based on normalized mutual information and brightness value correction; (S150) Non-rigid body matching applying tumor rigid constraints;
Abstract translation: 目的：提供一种应用肿瘤刚度约束和动态收缩增强乳房MR图像中每个组织的密度校正的非刚性配准方法和系统，以通过正则化对比度密度来精确地分类组织。 构成：通过去除乳房皮肤和胸部，自动分割乳房区域（S110）。 在分割的乳房区域中使用模糊C均值分组方法。 乳腺组织分为脂肪，乳腺，肿瘤和血管（S120）。 通过使用密度和形态学手术分割肿瘤（S130）。 在脂肪和乳腺中校正密度（S140）。 （附图标记）（AA）开始; （BB）结束; （S110）基于斜率和方向信息的乳房分割; （S120）使用模糊C均值分组法进行组织分类; （S130）基于亮度值和形状信息的肿瘤分割; （S140）基于归一化互信息和亮度值校正的刚体匹配; （S150）应用肿瘤刚性约束的非刚体匹配;
-
公开(公告)号：KR1020110077795A
公开(公告)日：2011-07-07
申请号：KR1020090134456
申请日：2009-12-30
Applicant: 서울여자대학교 산학협력단
IPC: A61B6/03
CPC classification number: A61B6/032 , A61B6/503 , A61B6/52 , G06T7/136 , G06T2207/10081
Abstract: PURPOSE: A device and method for detecting a boundary between an inner wall and an outer wall of a left ventricle are provided to set a proper brightness critical range at an image with a low contrast ratio by considering a brightness value distribution of a myocardium according to each data property. CONSTITUTION: A myocardium inner wall extracting unit(210) computes a seed position from a CT(Computed Tomography) image of a heart and extracts the boundary of an initial myocardium inner wall by using a radiation tracking method. A myocardium outer wall extracting unit(220) extracts the boundary of a myocardium outer wall from the CT image of the heart. A boundary correcting unit(230) corrects a boundary between the inner wall and the outer wall of the myocardium.
Abstract translation: 目的：提供一种用于检测左心室内壁和外壁之间边界的装置和方法，以通过考虑根据心脏的心肌的亮度值分布来设置具有低对比度的图像的适当的亮度临界范围 每个数据属性。 构成：心肌内壁提取单元（210）从心脏的CT（计算机断层摄影）图像计算种子位置，并通过使用辐射跟踪方法提取初始心肌内壁的边界。 心肌外壁提取单元（220）从心脏的CT图像提取心肌外壁的边界。 边界校正单元（230）校正心肌的内壁和外壁之间的边界。
-
Patent Agency Ranking
公开(公告)号：KR1020110077740A
公开(公告)日：2011-07-07
申请号：KR1020090134389
申请日：2009-12-30
Applicant: 서울여자대학교 산학협력단
IPC: A61B6/03
CPC classification number: A61B6/032 , A61B6/504 , A61B6/52 , G06T2207/10081
Abstract: PURPOSE: An apparatus and method for automatically extracting a blood vessel and calcification are provided to accurately extract a thin blood vessel without damaging peripheral tissues by removing a complex bone structure on a CTA(Computed Tomography Angiography) image. CONSTITUTION: A partition unit(110) partitions the lower limbs into a preset number of areas by using bone connection information. A matching unit(120) matches the partitioned areas with a rigid body using distance map of CT(Computed Tomography) and CTA. A blood vessel extracting unit(130) extracts the blood vessel by using brightness value information and blood masking. A noise removing unit(140) removes noises due to rigid body matching errors by using a blood vessel based post-processing method. A calcium extracting unit(150) extracts calcium by tracking the extracted blood vessel.
Abstract translation: 目的：提供一种用于自动提取血管和钙化的装置和方法，以通过去除CTA（计算机断层扫描血管造影）图像上的复杂骨结构来精确地提取薄血管而不损伤外周组织。 构成：分隔单元（110）通过使用骨骼连接信息将下肢分成预定数量的区域。 匹配单元（120）使用CT（计算机断层扫描）和CTA的距离图匹配具有刚体的分割区域。 血管提取单元（130）通过使用亮度值信息和血液掩蔽来提取血管。 噪声去除单元（140）通过使用基于血管的后处理方法来消除由于刚体匹配误差引起的噪声。 钙提取单元（150）通过跟踪提取的血管来提取钙。
-
公开(公告)号：KR101028365B1
公开(公告)日：2011-04-11
申请号：KR1020090009063
申请日：2009-02-04
Applicant: 서울여자대학교 산학협력단 , 서울대학교산학협력단
IPC: A61B6/02
Abstract: 본 발명은 연속적으로 촬영된 컴퓨터 단층촬영(Computed Tomography, CT) 영상의 각 슬라이스를 정합함으로써 연속 CT 영상 내의 폐결절 (Pulmonary Nodule)을 추적 관찰하는 방법 및 그 장치를 제공할 수 있다. 본 발명은 폐를 포함하는 최적경계볼륨의 중심 좌표를 기반으로 기준 영상에 대한 추적 영상의 위치를 보정하는 단계, 기준 영상 및 추적 영상 간의 관상(coronal) 최대강도 투사영상 간의 비교 결과 및 시상(sagittal) 최대강도 투사영상 간의 비교 결과에 기반하여 상기 기준 영상 및 상기 추적 영상을 강체 정합하는 단계, 및 상기 기준 영상 및 상기 추적 영상 내의 폐결절의 형태 정보 및 주변 지역 정보에 기반하여 상기 기준 영상 및 상기 추적 영상을 템플릿 정합하는 단계를 포함할 수 있다.
컴퓨터 단층촬영 영상, 폐결절, 관상 최대강도 투사, 시상 최대강도 투사, 강체 정합, 템플릿 매칭-
48.发明公开자기 공명 영상에서 밝기 값 분포 및 기울기 정보를 이용한 활성 형상 모델 기반 전립선 영상 자동 분할 장치 및 방법 有权基于激光成像模型的自动分割的装置和方法，使用强磁性分布和磁共振图像中的梯度信息
公开(公告)号：KR1020110018768A
公开(公告)日：2011-02-24
申请号：KR1020090076394
申请日：2009-08-18
Applicant: 서울여자대학교 산학협력단
IPC: A61B5/055
Abstract: PURPOSE: In the site in which the brightness value of the prostate gland similarly appears in the magnetic resonance imaging with the peripheral operations. In the magnetic resonance imaging, the brightness value distribution and gradient information are used. CONSTITUTION: In the magnetic resonance imaging, the activity shape model base prostate gland image automatic segmentation apparatus and the method using the brightness value distribution and gradient information include the falling stage(S310). The step of being created the activity shape model initialized from the magnetic resonance imaging(S330). The step of extracting the surface of the prostate gland domain it predicts the activity shape model based on the information which the activity shape model exhibits(S340).
Abstract translation: 目的：在前列腺的亮度值同样出现在与外围手术的磁共振成像中的位置。 在磁共振成像中，使用亮度值分布和梯度信息。 构成：在磁共振成像中，活动形状模型基础前列腺图像自动分割装置和使用亮度值分布和梯度信息的方法包括下降阶段（S310）。 创建从磁共振成像初始化的活动形状模型的步骤（S330）。 提取前列腺表面的步骤是根据活动形状模型展现的信息预测活动形状模型（S340）。
-
公开(公告)号：KR100933569B1
公开(公告)日：2009-12-29
申请号：KR1020080036783
申请日：2008-04-21
Applicant: 서울여자대학교 산학협력단
Abstract: PURPOSE: A method and an apparatus of diagnosing prostate cancer using a histopathology image and a magnetic resonance image are provided to improve recognition of a tumor area of a T2-weighted magnetic resonance image by dense correction. CONSTITUTION: A histopathology image correcting unit(11) generates and corrects the entire histopathology image. The histopathology image correcting unit includes a section engaging unit(112) and a resolution lowering unit(114). A magnetic resonance image correcting unit(13) corrects a T2-weighted magnetic resonance image. The magnetic resonance image correcting unit includes a bleeding area extracting unit(132) and a bleeding area substitution unit(134). A matching processing unit(15) includes the first matching unit(152), the second matching unit(154) and the third matching unit(156).
Abstract translation: 目的：提供一种使用组织病理学图像和磁共振图像诊断前列腺癌的方法和装置，以通过密集校正来改善对T2加权磁共振图像的肿瘤区域的识别。 构成：组织病理学图像校正单元（11）生成并校正整个组织病理学图像。 组织病理学图像校正单元包括区段接合单元（112）和分辨率降低单元（114）。 磁共振图像校正单元（13）校正T2加权磁共振图像。 磁共振图像校正单元包括出血区域提取单元（132）和出血区域置换单元（134）。 匹配处理单元（15）包括第一匹配单元（152），第二匹配单元（154）和第三匹配单元（156）。
-
公开(公告)号：KR102228087B1
公开(公告)日：2021-03-15
申请号：KR1020190075813
申请日：2019-06-25
Applicant: 서울여자대학교 산학협력단 , 사회복지법인 삼성생명공익재단
Abstract: 컴퓨터가수행하는의료영상에서의특정연골분할방법이제공된다. 상기방법은특정연골을포함하는의료영상을획득하는단계, 상기의료영상에포함된상기특정연골및 상기특정연골의주변에위치하는주변구조물을고려하여, 상기의료영상에대해 2차원분할네트워크를이용하여분할(segmentation)하는단계, 상기 2차원분할네트워크를이용하여분할된결과를기초로상기특정연골의관심볼륨영역(VOI; Volume Of Interest)을추출하고, 상기관심볼륨영역을기초로 3차원분할네트워크를이용하여분할(segmentation)하는단계를포함한다.
-
-
-
-
-
-
-
-
-
Search Results
61
records
(took 0.034 seconds)
