Estimating Retinal Sensitivity Using Optical Coherence Tomography With Deep-Learning Algorithms in Macular Telangiectasia Type 2
Estimating Retinal Sensitivity Using Optical Coherence Tomography With Deep-Learning Algorithms in Macular Telangiectasia Type 2
复制标题
DOI：
10.1001/jamanetworkopen.2018.8029
复制
发表时间：
2019-02-01
影响因子：
13.8
通讯作者：
Lee, Aaron Y.
中科院分区：
文献类型：
--
作者：
Kihara, Yuka;Heeren, Tjebo F. C.;Lee, Aaron Y.
重要性 目前使用的微视野测量是一种繁重的临床测试方式，用于测试视网膜敏感性，需要较长的测试时间和训练有素的技术人员。 目的 创建一个深度学习网络，可以直接从头开始估计结构的功能，以提供估计视网膜敏感性的正面高分辨率图。 设计、设置和参与者 使用 2016 年 1 月 1 日至 2017 年 11 月 30 日期间收集的数据进行横断面成像研究， 2 型黄斑毛细血管扩张症自然史观察和登记处 (MacTel) 对来自 2 个中心的 38 名确诊 MacTel 的参与者进行了评估。 主要结果和测量 与观察到的视网膜敏感性相比，估计的平均绝对误差。通过眼底控制视野检查（微视野检查）获得观察到的视网膜敏感性。视网膜敏感性的估计是通过深度学习模型进行的，该模型学习高分辨率光学相干断层扫描（OCT）扫描和微视野检查结果的叠加。这些预测被用来创建黄斑的高密度面部敏感度图。训练、验证和测试集在患者层面进行隔离。 结果 总共 2499 个微视野检查灵敏度被映射到 38 名患者 63 只眼睛的 1708 次 OCT B 扫描上（平均 [SD] 年龄，74.3 [9.7] 岁；15 名男性 [39.5%]）。我们的算法的训练示例数量为 67 899 个（数据增强后为 103 053 个），验证示例数量为 1695 个，测试示例数量为 1212 个。使用线性回归时的平均绝对误差结果为 4.51 dB（95% CI，4.36-4.65 dB），使用 LeNet 模型时的平均绝对误差结果为 3.66 dB（95% CI，3.53-3.78 dB）。使用 4990 万变量深度学习模型，验证和测试的视网膜灵敏度平均绝对误差为 3.36 dB（95% CI，3.25-3.48 dB）。相关性显示出高度一致（Pearson 相关性 r = 0.78）。通过配对 Wilcoxon 秩和检验，我们的模型显着优于这 2 个基线模型 (P
IMPORTANCE As currently used, microperimetry is a burdensome clinical testing modality for testing retinal sensitivity requiring long testing times and trained technicians.OBJECTIVE To create a deep-learning network that could directly estimate function from structure de novo to provide an en face high-resolution map of estimated retinal sensitivity.DESIGN, SETTING, AND PARTICIPANTS A cross-sectional imaging study using data collected between January 1, 2016, and November 30, 2017, from the Natural History Observation and Registry of macular telangiectasia type 2 (MacTel) evaluated 38 participants with confirmed MacTel from 2 centers.MAIN OUTCOMES AND MEASURES Mean absolute error of estimated compared with observed retinal sensitivity. Observed retinal sensitivity was obtained with fundus-controlled perimetry (microperimetry). Estimates of retinal sensitivity were made with deep-learning models that learned on superpositions of high-resolution optical coherence tomography (OCT) scans and microperimetry results. Those predictions were used to create high-density en face sensitivity maps of the macula. Training, validation, and test sets were segregated at the patient level.RESULTS A total of 2499 microperimetry sensitivities were mapped onto 1708 OCT B-scans from 63 eyes of 38 patients (mean [SD] age, 74.3 [9.7] years; 15 men [39.5%]). The numbers of examples for our algorithm were 67 899 (103 053 after data augmentation) for training, 1695 for validation, and 1212 for testing. Mean absolute error results were 4.51 dB (95% CI, 4.36-4.65 dB) when using linear regression and 3.66 dB (95% CI, 3.53-3.78 dB) when using the LeNet model. Using a 49.9 million-variable deep-learning model, a mean absolute error of 3.36 dB (95% CI, 3.25-3.48 dB) of retinal sensitivity for validation and test was achieved. Correlation showed a high degree of agreement (Pearson correlation r = 0.78). By paired Wilcoxon rank sum test, our model significantly outperformed these 2 baseline models (P