光学学报, 2019, 39 (7): 0715001, 网络出版: 2019-07-16 复制并引用该论文  本文已被引 1 次  被引通知

图 & 表

图 1. 直角坐标系下支架植入后的IVOCT图像和支架轮廓可由专家标记的4个支架角点表示过程。(a)单个放大的支架;(b)专家在单个支架标记4个角点;(c)基于4个标记角点得到支架轮廓图;(d)直角坐标系下支架植入后的IVOCT图像

Fig. 1. IVOCT images after BVS stenting in Cartesian coordinate system and process where strut contour can be represented from four labeled corners by expert. (a) Single enlarged strut; (b) four corners labeled on single strut by expert; (c) strut contour obtained from four labeled corners; (d) IVOCT image after BVS stenting in Cartesian coordinate system

下载图片 查看原文

图 2. 基于不同转换中心得到的两种IVOCT极坐标图像。(a)以直角坐标下图像的图像中心为转换中心的IVOCT极坐标图像;(b)以直角坐标下图像的管腔轮廓中心为转换中心的IVOCT极坐标图像

Fig. 2. IVOCT images shown in polar coordinates based on two different transformed centers. (a) IVOCT image in polar coordinate system using center of image in Cartesian coordinate system as transformed center; (b) IVOCT image in polar coordinate system using lumen-contour center of image in Cartesian coordinate system as transformed center

下载图片 查看原文

图 3. 选取正负训练样本

Fig. 3. Selection of positive and negative training samples

下载图片 查看原文

图 4. 基本哈尔特征模板及基于支架特征得到的扩展模板。(a)~(e)基本哈尔特征模板;(f)~(j)基于支架特征的扩展模板

Fig. 4. Basic prototypes with Haar-like features and extended prototypes combined with struts' features. (a)-(e) Basic prototypes with Haar-like features; (f)-(j) extended prototypes combined with struts' features

下载图片 查看原文

图 5. 支架4个角点检测流程图。(a)~(d)使用十字划分方法得到支架的4个角点流程图;(e)~(h)使用支架角度划分方法得到4个角点流程图;(i)支架角度获取方法

Fig. 5. Flow charts of four corners detection of struts. (a)-(d) Flow chart of four corners detection by cross-division method; (e)-(h) flow chart of four corners detection by strut-angle-division method; (i) method of getting strut angle

下载图片 查看原文

图 6. 支架分割结果对比图。(a)~(d)使用DP算法得到的支架分割结果;(e)~(h)使用本文中角点检测方法得到的支架分割结果

Fig. 6. Comparison of strut segmentation results. (a)-(d) Results of BVS strut segmentation using DP algorithm; (e)-(h) results of BVS strut segmentation using proposed algorithms

下载图片 查看原文

图 7. BVS支架3D重建模型图。(a)含有冠状动脉血管的BVS支架3D重建图;(b)去除冠状动脉血管的BVS支架3D重建图

Fig. 7. 3D reconstruction models of BVS struts. (a) 3D reconstruction model of BVS strut with coronary artery; (b) 3D reconstruction model of BVS strut without coronary artery

下载图片 查看原文

表 1支架检测与分割结果

Table1. Results of strut detection and segmentation

Data set No. F No. GT Ecp /μm Segmentation(Dice) Method in Ref.[10] Proposed No.1 81 691 24.77±15.08 0.80 0.90 No.2 119 928 27.88±20.39 0.79 0.89 No.3 118 1172 31.45±20.48 0.80 0.87 No.4 78 604 25.27±18.45 0.82 0.89 No.5 147 1188 30.68±19.84 0.82 0.88 No.6 86 635 29.80±19.54 0.80 0.89 No.7 76 603 27.84±19.36 0.79 0.89 No.8 150 1240 30.23±19.96 0.80 0.87 Average - - 28.99±19.39 0.80 0.88 No. F: number of frames evaluated; No. GT: number of ground truth

查看原文

姚林林, 金琴花, 荆晶, 陈韵岱, 曹一挥, 李嘉男, 朱锐. 基于角点检测的可降解支架轮廓分割算法[J]. 光学学报, 2019, 39(7): 0715001. Linlin Yao, Qinhua Jin, Jing Jing, Yundai Chen, Yihui Cao, Jianan Li, Rui Zhu. Corner Detection-Based Segmentation Algorithm of Bioresorbable Vascular Scaffold Strut Contours[J]. Acta Optica Sinica, 2019, 39(7): 0715001.