文献速递青光眼神经眼科

光学相干断层扫描图像中视神经乳头结构的自动分割算法评估

Evaluation of Automated Segmentation Algorithms for Optic Nerve Head Structures in Optical Coherence Tomography Images.
PubMed 文献 · Invest Ophthalmol Vis Sci · 2018-01-01

本研究在29例青光眼患者及疑似者的51只眼中,比较观察者与自动算法在光学相干断层扫描图像上标记视神经乳头结构(布鲁赫膜开口BMO和筛板前部LC)的一致性。结果显示自动算法在BMO宽度和LC位置上与观察者标记无显著差异,重复分析变异性小,BMO可作为算法分析的实用参考平面。

要点

  • 自动算法与观察者在BMO宽度和LC位置上标记无显著差异。
  • 算法重复分析BMO面积和筛板前部深度变异小。
  • 水平视神经乳头区域筛板前部深度变异性较大。

核心结论

  • 自动识别视神经乳头结构与观察者标记在BMO和筛板前部位置上相当。
  • BMO可作为算法分析的实用参考平面。

研究设计与样本

  • 在29例青光眼患者及疑似者的51只眼中获取24个放射扫描集。
  • 进行设盲的观察者内和观察者间比较,并比较观察者与自动算法的标记。
  • 对20只眼进行算法重复分析比较。

主要发现

  • 观察者内BMO宽度差异与零无显著差异(P ≥ 0.32),LC位置差异小于1%(P = 0.04)。
  • 观察者间BMO宽度差异为0.36%(P = 0.63),LC位置平均差异为14.74 μm(P = 0.004),为典型筛板前部深度的3%。
  • 观察者与算法间BMO宽度差异为1.85%(P = 0.23),LC位置平均差异无统计学意义(3.77 μm,P = 0.77)。
  • 算法重复分析BMO面积平均差异为0.38%(P = 0.47),筛板前部深度平均差异为0.54 ± 0.72%。
  • 水平视神经乳头区域筛板前部深度变异性较大,部分离群图像未被有效标记。

关键数值与推荐

  • 观察者内BMO宽度差异P ≥ 0.32,LC位置差异P = 0.04。
  • 观察者间BMO宽度差异0.36%(P = 0.63),LC位置差异14.74 μm(P = 0.004)。
  • 观察者与算法间BMO宽度差异1.85%(P = 0.23),LC位置差异3.77 μm(P = 0.77)。
  • 算法重复分析BMO面积差异0.38%(P = 0.47),筛板前部深度差异0.54 ± 0.72%。

临床怎么用

  • 自动算法可用于视神经乳头BMO和筛板前部位置的定量分析。
  • BMO可作为算法分析的参考平面。
  • 需注意水平视神经乳头区域筛板前部深度变异性较大。

局限性

  • 部分个体离群图像未被观察者或算法有效标记。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

PURPOSE: To compare the identification of optic nerve head (ONH) structures in optical coherence tomography images by observers and automated algorithms. METHODS: ONH images in 24 radial scan sets by optical coherence tomography were obtained in 51 eyes of 29 glaucoma patients and suspects. Masked intraobserver and interobserver comparisons were made of marked endpoints of Bruch's membrane opening (BMO) and the anterior lamina cribrosa (LC). BMO and LC positional markings were compared between observer and automated algorithm. Repeated analysis on 20 eyes by the algorithm was compared. Regional ONH data were derived from the algorithms. RESULTS: Intraobserver difference in BMO width was not significantly different from zero (P ≥ 0.32) and the difference in LC position was less than 1% different (P = 0.04). Interobserver were slightly larger than intraobserver differences, but interobserver BMO width difference was 0.36% (P = 0.63). Mean interobserver difference in LC position was 14.74 μm (P = 0.004), 3% of the typical anterior lamina depth (ALD). Between observer and algorithm, BMO width differed by 1.85% (P = 0.23) and mean LC position was not significantly different (3.77 μm, P = 0.77). Repeat algorithmic analysis had a mean difference in BMO area of 0.38% (P = 0.47) and mean ALD difference of 0.54 ± 0.72%. Regional ALD had greater variability in the horizontal ONH regions. Some individual outlier images were not validly marked by either observers or algorithm. CONCLUSIONS: Automated identification of ONH structures is comparable to observer markings for BMO and anterior LC position, making BMO a practical reference plane for algorithmic analysis.

来源:Invest Ophthalmol Vis Sci,查看原文。版权归原作者所有。

引用

  1. Duan XJ, Jefferys JL, Quigley HA, et al. Evaluation of Automated Segmentation Algorithms for Optic Nerve Head Structures in Optical Coherence Tomography Images. Invest Ophthalmol Vis Sci. 2018. PMID: 30073355. doi:10.1167/iovs.18-24469

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。