Investigating Determinants and Evaluating Deep Learning Training Approaches for Visual Acuity in Foveal Hypoplasia.
Investigating Determinants and Evaluating Deep Learning Training Approaches for Visual Acuity in Foveal Hypoplasia.
复制标题
DOI：
10.1016/j.xops.2022.100225
复制
发表时间：
2023-03
影响因子：
--
通讯作者：
Hufnagel, Robert B.
中科院分区：
文献类型：
--
作者：
Malechka, Volha V.;Duong, Dat;Bordonada, Keyla D.;Turriff, Amy;Blain, Delphine;Murphy, Elizabeth;Introne, Wendy J.;Gochuico, Bernadette R.;Adams, David R.;Zein, Wadih M.;Brooks, Brian P.;Huryn, Laryssa A.;Solomon, Benjamin D.;Hufnagel, Robert B.
描述一组患有中心凹发育不全 (FH) 的个体的中心凹结构和视觉功能之间的关系，并使用深度学习分类器估计 FH 等级和视力。回顾性队列研究和实验研究。 2004 年至 2018 年，国家眼科研究所总共对 201 名 FH 患者进行了评估。分析了中心凹 OCT 扫描的结构成分和相应的临床数据，以评估其对视力的贡献。为了自动化 FH 评分和视敏度相关性，我们评估了以下 3 个用于训练神经网络预测器的输入：(1) OCT 扫描，(2) OCT 扫描和元数据，以及 (3) 从生成对抗网络创建的真实 OCT 扫描和假 OCT 扫描。视力结果与决定因素（例如中心凹形态、眼球震颤和屈光不正）之间的关系。 OCT 成像时受试者的平均年龄为 24.4 岁（范围：1-73 岁；标准差 = 18.25 岁）。平均最佳矫正视力（n = 398 只眼睛）相当于最小分辨率角度 (LogMAR) 值 0.75 (Snellen 20/115) 的对数。等效球面屈光不正 (SER) 范围为 -20.25 屈光度 (D) 至 +13.63 D，中位数为 +0.50 D。眼球震颤的存在和高 LogMAR 值显示出统计上显着的关系 (P < 0.0001)。 SER 值远离 plano 的参与者表现出较高的 LogMAR 值（n = 382 只眼睛）。随着 FH 级别的升高，出现眼球震颤的患者比例也随之增加。与 1 级（范围 -8.88 D 至 +8.50 D）相比，4 级（范围 -20.25 D 至 +13.00 D）SER 的变异性具有统计学显着性（P < 0.0001）。我们的神经网络预测器可靠地估计了 37 人测试队列（98 次 OCT 扫描）的 FH 分级和视力（与真实值的相关性分别 > 0.85 和 > 0.70）。使用元数据和假 OCT 扫描在真实 OCT 扫描上训练预测器，比仅在真实 OCT 扫描上训练的模型提高了准确性。眼球震颤和中心凹解剖结构影响 FH 患者的视觉结果，计算算法可以可靠地估计 FH 分级和视力。
To describe the relationships between foveal structure and visual function in a cohort of individuals with foveal hypoplasia (FH) and to estimate FH grade and visual acuity using a deep learning classifier. Retrospective cohort study and experimental study. A total of 201 patients with FH were evaluated at the National Eye Institute from 2004 to 2018. Structural components of foveal OCT scans and corresponding clinical data were analyzed to assess their contributions to visual acuity. To automate FH scoring and visual acuity correlations, we evaluated the following 3 inputs for training a neural network predictor: (1) OCT scans, (2) OCT scans and metadata, and (3) real OCT scans and fake OCT scans created from a generative adversarial network. The relationships between visual acuity outcomes and determinants, such as foveal morphology, nystagmus, and refractive error. The mean subject age was 24.4 years (range, 1–73 years; standard deviation = 18.25 years) at the time of OCT imaging. The mean best-corrected visual acuity (n = 398 eyes) was equivalent to a logarithm of the minimal angle of resolution (LogMAR) value of 0.75 (Snellen 20/115). Spherical equivalent refractive error (SER) ranged from −20.25 diopters (D) to +13.63 D with a median of +0.50 D. The presence of nystagmus and a high-LogMAR value showed a statistically significant relationship (P < 0.0001). The participants whose SER values were farther from plano demonstrated higher LogMAR values (n = 382 eyes). The proportion of patients with nystagmus increased with a higher FH grade. Variability in SER with grade 4 (range, −20.25 D to +13.00 D) compared with grade 1 (range, −8.88 D to +8.50 D) was statistically significant (P < 0.0001). Our neural network predictors reliably estimated the FH grading and visual acuity (correlation to true value > 0.85 and > 0.70, respectively) for a test cohort of 37 individuals (98 OCT scans). Training the predictor on real OCT scans with metadata and fake OCT scans improved the accuracy over the model trained on real OCT scans alone. Nystagmus and foveal anatomy impact visual outcomes in patients with FH, and computational algorithms reliably estimate FH grading and visual acuity.
登录
查看更多内容
影响因子：
3.7
作者：
通讯作者：
--
影响因子：
--
作者：
Chen, Jimmy S.;Coyner, Aaron S.;Chan, R. V. Paul;Hartnett, M. Elizabeth;Moshfeghi, Darius M.;Owen, Leah A.;Kalpathy-Cramer, Jayashree;Chiang, Michael F.;Campbell, J. Peter
通讯作者：
Campbell, J. Peter
影响因子：
--
作者：
Langlo CS;Trotter A;Reddi HV;Schilter KF;Tyler RC;Udani R;Neitz M;Carroll J;Connor TB
通讯作者：
Connor TB
影响因子：
--
作者：
Gale MJ;Titus HE;Harman GA;Alabduljalil T;Dennis A;Wilson JL;Koeller DM;Finanger E;Blasco PA;Chiang PW;Karr DJ;Yang P
通讯作者：
Yang P
影响因子：
4.4
作者：
Lee, Helena;Purohit, Ravi;Gottlob, Irene
通讯作者：
Gottlob, Irene