Belghith Akram, Bowd Christopher, Weinreb Robert N, Zangwill Linda M
Annu Int Conf IEEE Eng Med Biol Soc. 2014;2014:3869-72. doi: 10.1109/EMBC.2014.6944468.
Glaucoma is a chronic neurodegenerative disease characterized by loss of retinal ganglion cells, resulting in distinctive changes in the optic nerve head (ONH) and retinal nerve fiber layer (RNFL). Important advances in technology for non-invasive imaging of the eye have been made providing quantitative tools to measure structural changes in ONH topography, a crucial step in diagnosing and monitoring glaucoma. 3D spectral domain optical coherence tomography (SD-OCT), an optical imaging technique, has been commonly used to discriminate glaucomatous from healthy subjects. In this paper, we present a new approach for locating the Bruch's membrane opening BMO and then estimating the optic disc size and rim area of 3D Spectralis SD-OCT images. To deal with the overlapping of the Bruch's membrane BM layer and the border tissue of Elschnig due to the poor image resolution, we propose the use of image deconvolution approach to separate these layers. To estimate the optic disc size and rim area, we propose the use of a new regression method based on the artificial neural network principal component analysis (ANN-PCA), which allows us to model irregularity in the BMO estimation due to scan shifts and/or poor image quality. The diagnostic accuracy of rim area, and rim to disc area ratio is compared to the diagnostic accuracy of global RNFL thickness measurements provided by two commercially available SD-OCT devices using receiver operating characteristic curve analyses.
青光眼是一种慢性神经退行性疾病，其特征是视网膜神经节细胞丢失，导致视神经乳头（ONH）和视网膜神经纤维层（RNFL）出现明显变化。眼部非侵入性成像技术取得了重要进展，提供了测量ONH地形结构变化的定量工具，这是诊断和监测青光眼的关键步骤。三维光谱域光学相干断层扫描（SD-OCT）是一种光学成像技术，常用于区分青光眼患者和健康受试者。在本文中，我们提出了一种新方法，用于定位布鲁赫膜开口（BMO），然后估计三维Spectralis SD-OCT图像的视盘大小和边缘面积。由于图像分辨率较差，布鲁赫膜（BM）层与埃尔施尼格边缘组织存在重叠，为解决这一问题，我们建议使用图像去卷积方法来分离这些层。为了估计视盘大小和边缘面积，我们建议使用一种基于人工神经网络主成分分析（ANN-PCA）的新回归方法，该方法使我们能够对由于扫描移位和/或图像质量差导致的BMO估计中的不规则性进行建模。使用接收器操作特征曲线分析，将边缘面积以及边缘与视盘面积比的诊断准确性与两种商用SD-OCT设备提供的全局RNFL厚度测量的诊断准确性进行比较。
