Assessment of frame-averaging algorithms in OCT image analysis.
Assessment of frame-averaging algorithms in OCT image analysis.
复制标题
DOI：
10.3928/23258160-20130313-09
复制
发表时间：
2013-03
期刊：
影响因子：
--
通讯作者：
Huang D
中科院分区：
文献类型：
--
作者：
Wu W;Tan O;Pappuru RR;Duan H;Huang D
评估光学相干断层扫描的帧配准和平均算法。对正常眼和青光眼眼（各n=20）进行成像。通过比较噪声方差（NV）、边缘扩展（SE）和对比噪声比（CNR）来测量客观差异。还评价了主观图像质量。两种帧平均算法（FA400和FA407）的NV和CNR较单帧算法好，但SE较差（p<0.01）。此外，这两种算法提供了更好的结构边界的主观评估比单一的图像（p<0.001）。与FA400相比，算法FA407具有显著更低的SE和更好的ILM可视化。帧平均显着抑制斑点噪声，并增加了视网膜结构的可见性。然而，不完美的图像配准引起的边缘模糊，可以检测到的SE参数。在帧平均算法中，较高的CNR和较低的NV表示更好的噪声抑制，但SE在比较算法之间的边缘保护方面最敏感。
To evaluate frame registration and averaging algorithms for optical coherence tomography. Normal and glaucoma eyes (n=20 each) were imaged. Objective differences were measured by comparing noise variance (NV), spread of edge (SE), and contrast-to-noise ratio (CNR). Subjective image quality was also evaluated. Two frame-averaging algorithms (FA400 and FA407) had better NV and CNR, but worse SE than did single frames (p<0.01). Additionally, both algorithms provided better subjective assessments of structure boundaries than did single images (p<0.001). Algorithm FA407 had significantly lower SE and better ILM visualization than did FA400. Frame-averaging significantly suppressed speckle noise and increased the visibility of retinal structures. However, imperfect image registration caused edge blurring that could be detected by the SE parameter. In frame-averaging algorithms, higher CNR and lower NV indicated better noise suppression, but SE was most sensitive in comparing edge preservation between algorithms.
登录
查看更多内容
影响因子：
4.4
作者：
Geissler, Alexander;Gartus, Andreas;Barth, Markus
通讯作者：
Barth, Markus
DOI：
10.1006/jvci.2001.0473
发表时间：
2001-09-01
影响因子：
2.6
作者：
Sanz, I;Calvo, ML;Lakshminarayanan, V
通讯作者：
Lakshminarayanan, V
DOI：
10.1109/t-su.1978.30978
发表时间：
1978-01-01
期刊：
IEEE TRANSACTIONS ON SONICS AND ULTRASONICS
影响因子：
--
作者：
BURCKHARDT, CB
通讯作者：
BURCKHARDT, CB
影响因子：
10.6
作者：
Lewis, A. S.;Knowles, G.
通讯作者：
Knowles, G.
影响因子：
3.5
作者：
TAPIOVAARA, MJ;WAGNER, RF
通讯作者：
WAGNER, RF