Forecasting future Humphrey Visual Fields using deep learning
Forecasting future Humphrey Visual Fields using deep learning
复制标题
DOI：
10.1371/journal.pone.0214875
复制
发表时间：
2019-04-05
期刊：
影响因子：
3.7
通讯作者：
Lee, Aaron Y.
中科院分区：
文献类型：
--
作者：
Wen, Joanne C.;Lee, Cecilia S.;Lee, Aaron Y.
PurposeTo确定是否可以训练深度学习网络来预测未来的24-2 Humphrey Visual Fields（HVFs）. Methods 1998年至2018年连续24-2 HVFs的所有数据点都是从大学数据库中提取的。使用具有保持测试集的十重交叉验证来开发模型开发的三个主要阶段：模型架构选择，数据集组合选择和具有迁移学习的时间间隔模型训练，以训练能够生成逐点视野预测的深度学习人工神经网络。逐点平均绝对误差（PMAE）和平均偏差（MD）之间的预测和实际的未来HVF的差异calculated.ResultsMore超过170万视野点提取到百分之一分贝从32,443 24-2 HVF。选择了具有2000万个可训练参数的最佳模型CascadeNet 5。测试集的整体逐点PMAE为2.47 dB（95% CI：2.45 dB至2.48 dB），深度学习显示出比线性模型有统计学显著改善。这100个经过充分训练的模型成功预测了未来长达5.5年的昏迷眼的未来HVF，预测和实际未来HVF的MD之间的相关性为0.92，平均差异为0.41 dB。深度学习网络显示出不仅能够学习时空HVF变化，而且能够生成对未来长达5.5年的HVF的预测，只有一个HVF。
PurposeTo determine if deep learning networks could be trained to forecast future 24-2 Humphrey Visual Fields (HVFs).MethodsAll data points from consecutive 24-2 HVFs from 1998 to 2018 were extracted from a university database. Ten-fold cross validation with a held out test set was used to develop the three main phases of model development: model architecture selection, dataset combination selection, and time-interval model training with transfer learning, to train a deep learning artificial neural network capable of generating a point-wise visual field prediction. The pointwise mean absolute error (PMAE) and difference in Mean Deviation (MD) between predicted and actual future HVF were calculated.ResultsMore than 1.7 million perimetry points were extracted to the hundredth decibel from 32,443 24-2 HVFs. The best performing model with 20 million trainable parameters, CascadeNet5, was selected. The overall point-wise PMAE for the test set was 2.47 dB (95% CI: 2.45 dB to 2.48 dB), and deep learning showed a statistically significant improvement over linear models. The 100 fully trained models successfully predicted future HVFs in glaucomatous eyes up to 5.5 years in the future with a correlation of 0.92 between the MD of predicted and actual future HVF and an average difference of 0.41 dB.ConclusionsUsing unfiltered real-world datasets, deep learning networks show the ability to not only learn spatio-temporal HVF changes but also to generate predictions for future HVFs up to 5.5 years, given only a single HVF.