Contour is an important pattern descriptor in image processing and particularly in region description, registration and length estimation. In many applications where contour is used, a good segmentation and an efficie...Contour is an important pattern descriptor in image processing and particularly in region description, registration and length estimation. In many applications where contour is used, a good segmentation and an efficient smoothing method are needed. In X-ray images, such as mammograms, where object edge is not clearly discernible, estimating the object’s contour may yield substantial shift along the boundary due to noise or segmentation drawbacks. An appropriate smoothing is therefore required to reduce these effects. In this paper, an approach based on local adaptive threshold segmentation to extract contour and a new smoothing approach founded on Fourier descriptors are introduced. The experimental results of extraction obtained from a set of mammograms and compared with the breast regions delineated by radiologists yielded a percent overlap area of 98.7% ± 0.9% with false positive and negative rates of 0.36 ± 0.74 and 0.93 ± 0.44 respectively. The proposed method was tested on a set of images and improved the accuracy, leading to an average error of less than one pixel.展开更多
Background: Radiotherapy (RT) techniques after Conservative Breast Surgery (CBS) vary. Three Dimension (3D) planning allows for better plan optimization compared to 2 Dimension (2D) plans and also allowing for creatin...Background: Radiotherapy (RT) techniques after Conservative Breast Surgery (CBS) vary. Three Dimension (3D) planning allows for better plan optimization compared to 2 Dimension (2D) plans and also allowing for creating Dose Volume Histograms (DVHs) for both Planning Target Volume (PTV) and Organs at Risk (OAR). Patients and Methods: Twenty consecutive patients with CBS planned for whole breast and supraclavicular (SCV) RT at the National Cancer Institute (NCI), Egypt between January and June 2016 were included in this study. All patients were planned clinically in 2D fashion with no more than 2 cm of ipsilateral lung allowed in the tangential fields “Limited 2D” (Limit-2D) then Target and OAR volumes were drawn according to the Radiation Therapy Oncology Group (RTOG) guidelines and 3D plans and a central slice PTV-based 2D plan, “Modified 2D” (Mod-2D), were performed in the same Computerized Tomography (CT) slices for each patient. Mono-Iso-Centeric technique (MIT) was used in 3D plans. DVH parameters were used to compare the three plans. Results: In 3D plans, compared to Limit-2D, coverage improved for the intact breast (V95% = 95% versus (Vs) 69%, p = 0.036) and SCVPTV (V90% = 90% Vs 65%, p = 0.01). The breast and SCV V 107%, V112% and Dmax were better with 3D plan however not statistical significant (NS). Junctional hot spots were 120% and 107% in the Limit-2D and 3D plans respectively (p = 0.04). The dose to the heart, mean (333 Vs 491 cGy), V10 (5% Vs 10%) and V20 (3% Vs 7%), Ipsilateral lung V20 (19% Vs 26%), and contra lateral breast D-max (205 Vs 462 cGy) were higher in 3D plans however NS, and the dose to the cord was the same. Comparison between 3D and Mod-2D showed better OAR sparing with 3D with mean heart dose (491 cGy Vs 782 cGy, p = 0.025) and Ipsilateral lung V20 (26% Vs 32%, p = 0.07% with statistically comparable target coverage. Conclusion: This study demonstrated that application of 3D plan using MIT improves coverage of breast and SCVPTVs with minimizing hot spot at the junctional area if compared with Limit-2D plans with comparable dose to OAR. When compared with Mod-2D plans, 3D plans not only had better target coverage but also better sparing of OAR, the latter was statistically significant.展开更多
基于区域的几何活动轮廓(Chan-Vese,CV)模型是乳腺超声图像中常用的一种分割算法。但传统的CV模型不能满足乳腺超声图像分割精度高、速度快的要求。因此,文章提出了一种基于指数加权平均比率(Ratio of Exponential Weighted Averages,RO...基于区域的几何活动轮廓(Chan-Vese,CV)模型是乳腺超声图像中常用的一种分割算法。但传统的CV模型不能满足乳腺超声图像分割精度高、速度快的要求。因此,文章提出了一种基于指数加权平均比率(Ratio of Exponential Weighted Averages,ROEWA)算子改进的CV模型,用于乳腺超声图像中病灶区域的分割。首先,计算乳腺超声图像的ROEWA算子。其次,基于图像的ROEWA算子构建边缘指示函数,用于代替CV模型中的Dirac项。最后,去除平滑项,从而提高曲线演化的速度。实验结果表明,文章提出的算法不仅能提高分割的精度,而且能显著提高分割的速度。展开更多
文摘Contour is an important pattern descriptor in image processing and particularly in region description, registration and length estimation. In many applications where contour is used, a good segmentation and an efficient smoothing method are needed. In X-ray images, such as mammograms, where object edge is not clearly discernible, estimating the object’s contour may yield substantial shift along the boundary due to noise or segmentation drawbacks. An appropriate smoothing is therefore required to reduce these effects. In this paper, an approach based on local adaptive threshold segmentation to extract contour and a new smoothing approach founded on Fourier descriptors are introduced. The experimental results of extraction obtained from a set of mammograms and compared with the breast regions delineated by radiologists yielded a percent overlap area of 98.7% ± 0.9% with false positive and negative rates of 0.36 ± 0.74 and 0.93 ± 0.44 respectively. The proposed method was tested on a set of images and improved the accuracy, leading to an average error of less than one pixel.
文摘Background: Radiotherapy (RT) techniques after Conservative Breast Surgery (CBS) vary. Three Dimension (3D) planning allows for better plan optimization compared to 2 Dimension (2D) plans and also allowing for creating Dose Volume Histograms (DVHs) for both Planning Target Volume (PTV) and Organs at Risk (OAR). Patients and Methods: Twenty consecutive patients with CBS planned for whole breast and supraclavicular (SCV) RT at the National Cancer Institute (NCI), Egypt between January and June 2016 were included in this study. All patients were planned clinically in 2D fashion with no more than 2 cm of ipsilateral lung allowed in the tangential fields “Limited 2D” (Limit-2D) then Target and OAR volumes were drawn according to the Radiation Therapy Oncology Group (RTOG) guidelines and 3D plans and a central slice PTV-based 2D plan, “Modified 2D” (Mod-2D), were performed in the same Computerized Tomography (CT) slices for each patient. Mono-Iso-Centeric technique (MIT) was used in 3D plans. DVH parameters were used to compare the three plans. Results: In 3D plans, compared to Limit-2D, coverage improved for the intact breast (V95% = 95% versus (Vs) 69%, p = 0.036) and SCVPTV (V90% = 90% Vs 65%, p = 0.01). The breast and SCV V 107%, V112% and Dmax were better with 3D plan however not statistical significant (NS). Junctional hot spots were 120% and 107% in the Limit-2D and 3D plans respectively (p = 0.04). The dose to the heart, mean (333 Vs 491 cGy), V10 (5% Vs 10%) and V20 (3% Vs 7%), Ipsilateral lung V20 (19% Vs 26%), and contra lateral breast D-max (205 Vs 462 cGy) were higher in 3D plans however NS, and the dose to the cord was the same. Comparison between 3D and Mod-2D showed better OAR sparing with 3D with mean heart dose (491 cGy Vs 782 cGy, p = 0.025) and Ipsilateral lung V20 (26% Vs 32%, p = 0.07% with statistically comparable target coverage. Conclusion: This study demonstrated that application of 3D plan using MIT improves coverage of breast and SCVPTVs with minimizing hot spot at the junctional area if compared with Limit-2D plans with comparable dose to OAR. When compared with Mod-2D plans, 3D plans not only had better target coverage but also better sparing of OAR, the latter was statistically significant.
文摘基于区域的几何活动轮廓(Chan-Vese,CV)模型是乳腺超声图像中常用的一种分割算法。但传统的CV模型不能满足乳腺超声图像分割精度高、速度快的要求。因此,文章提出了一种基于指数加权平均比率(Ratio of Exponential Weighted Averages,ROEWA)算子改进的CV模型,用于乳腺超声图像中病灶区域的分割。首先,计算乳腺超声图像的ROEWA算子。其次,基于图像的ROEWA算子构建边缘指示函数,用于代替CV模型中的Dirac项。最后,去除平滑项,从而提高曲线演化的速度。实验结果表明,文章提出的算法不仅能提高分割的精度,而且能显著提高分割的速度。
