Haeker Mona, Abràmoff Michael D, Wu Xiaodong, Kardon Randy, Sonka Milan
Department of Electrical & Computer Engineering, The University of Iowa, Iowa City, IA 52242, USA.
Med Image Comput Comput Assist Interv. 2007;10(Pt 1):244-51. doi: 10.1007/978-3-540-75757-3_30.
An optimal 3-D graph search approach designed for simultaneous multiple surface detection is extended to allow for varying smoothness and surface interaction constraints instead of the traditionally used constant constraints. We apply the method to the intraretinal layer segmentation of 24 3-D optical coherence tomography (OCT) images, learning the constraints from examples in a leave-one-subject-out fashion. Introducing the varying constraints decreased the mean unsigned border positioning errors (mean error of 7.3 +/- 3.7 microm using varying constraints compared to 8.3 +/- 4.9 microm using constant constraints and 8.2 +/- 3.5 microm for the inter-observer variability).
一种为同时进行多表面检测而设计的最优三维图形搜索方法得到了扩展，以允许采用变化的平滑度和表面相互作用约束，而非传统使用的恒定约束。我们将该方法应用于24幅三维光学相干断层扫描（OCT）图像的视网膜内部分层，以留一法从示例中学习约束。引入变化的约束降低了平均无符号边界定位误差（使用变化约束时平均误差为7.3±3.7微米，使用恒定约束时为8.3±4.9微米，观察者间变异性为8.2±3.5微米）。
