| 姚炜,蒋栋,赵宝珍,陈宁宁,钱国正.星形算法对乳头肌短轴切面左心室内膜全自动勾画的初步研究[J].中国医学影像技术,2005,21(7):1104~1107 |
| 星形算法对乳头肌短轴切面左心室内膜全自动勾画的初步研究 |
| Initial study of automatic left ventricular endocardial boundary detection in papillary short axis view using star algorithm |
| 投稿时间：2004-12-10 修订日期：2005-04-06 |
| DOI： |
| 中文关键词: 星形算法 全自动勾画 心内膜边界 超声心动描记术 |
| 英文关键词:Star algorithm Automatic detection Endocardial boundary Echocardiography |
| 基金项目: |
|
| 摘要点击次数: 3058 |
| 全文下载次数: 734 |
| 中文摘要: |
| 目的 初步测试星形算法的边界形状适应能力。方法 选择两乳头肌间连线中点作为形心,然后从形心出发向四周放射形探测心内膜,并且在勾画边界的代价函数中加入了反映形状特征的限制条件。结果 用六组具有典型左心室内膜边界形状的舒张末期和收缩末期乳头肌短轴切面图像对星形算法的全自动勾画进行了测试,并用手工勾画的心内膜边界作为参照。结果有一组勾画失败,在其余的五组图像中星形算法与手工勾画的心内膜相比,其勾画边界的平均半径误差率为21.80％。结论 星形算法具有较强的边界形状适应能力。 |
| 英文摘要: |
| Objective To testify the shape change adaptability of a new automatic echocardial endocardium detector called star algorithm. Methods The mid-point of the two papillary muscles was selected as the shape center, from which star algorithm detected left ventricular endocardium radically. And the cost function, which was used to calculate the endocardial boundary, was added by conditions reflecting shape features. Results Star algorithm was performed on six pairs of end-diastolic and end-systolic mid-papillary short axis images of left ventricle with various endocardial boundary shapes, of which the Results were compared with those of manually traced boundaries. Compared with the manually traced boundary, except for the failure in one image pair, the mean error radial length rate of star algorithm detected boundary was 21.80%. Conclusion The star algorithm has robust shape change adaptability in echocardial endocardium detection. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
|
|
|
