文献速递青光眼眼底病与葡萄膜炎

深度学习模型基于OCT预测青光眼中心10°视野

Deep learning model to predict visual field in central 10° from optical coherence tomography measurement in glaucoma.
PubMed 文献 · Br J Ophthalmol · 2020-06-27

本研究训练并验证深度学习模型,从SD-OCT厚度预测青光眼患者中心10°视野阈值。多中心横断面训练集纳入591眼,外部测试集160眼,CNN预测全视野平均阈值的绝对误差为2.84±2.98 dB,显著优于SVM和MLR(p<0.001),全视野平均阈值R²为0.74。

要点

  • CNN预测全视野平均阈值绝对误差2.84±2.98 dB,优于SVM和MLR(p<0.001)。
  • 逐点平均绝对误差CNN为5.47±3.05 dB,全视野平均阈值R²为0.74。
  • 模型基于黄斑三层厚度预测HFA 10-2视野阈值,外部测试集验证性能。

核心结论

  • 深度学习模型从SD-OCT厚度预测HFA 10-2视野阈值具有较好准确性。

研究设计与样本

  • 多中心横断面研究,训练集591眼(347例开角型青光眼或正常受试者),外部测试集160眼(131例开角型青光眼患者)。

主要发现

  • CNN预测全视野平均阈值绝对误差2.84±2.98 dB,显著小于SVM(5.65±5.12 dB)和MLR(6.96±5.38 dB)(均p<0.001)。
  • CNN逐点平均绝对误差5.47±3.05 dB,显著小于SVM(7.96±4.63 dB)和MLR(11.71±4.15 dB)(均p<0.001)。
  • CNN全视野平均阈值R²为0.74,68个点总体R²为0.44±0.24。

关键数值与推荐

  • CNN预测全视野平均阈值绝对误差2.84±2.98 dB,R²为0.74。
  • CNN逐点平均绝对误差5.47±3.05 dB,68点总体R²为0.44±0.24。
  • CNN性能显著优于SVM和MLR(均p<0.001)。

临床怎么用

  • 可辅助从SD-OCT黄斑层厚度预测中心10°视野阈值,减少视野检查负担。

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

BACKGROUND/AIM: To train and validate the prediction performance of the deep learning (DL) model to predict visual field (VF) in central 10° from spectral domain optical coherence tomography (SD-OCT). METHODS: This multicentre, cross-sectional study included paired Humphrey field analyser (HFA) 10-2 VF and SD-OCT measurements from 591 eyes of 347 patients with open-angle glaucoma (OAG) or normal subjects for the training data set. We trained a convolutional neural network (CNN) for predicting VF threshold (TH) sensitivity values from the thickness of the three macular layers: retinal nerve fibre layer, ganglion cell layer+inner plexiform layer and outer segment+retinal pigment epithelium. We implemented pattern-based regularisation on top of CNN to avoid overfitting. Using an external testing data set of 160 eyes of 131 patients with OAG, the prediction performance (absolute error (AE) and R2 between predicted and actual TH values) was calculated for (1) mean TH in whole VF and (2) each TH of 68 points. For comparison, we trained support vector machine (SVM) and multiple linear regression (MLR). RESULTS: AE of whole VF with CNN was 2.84±2.98 (mean±SD) dB, significantly smaller than those with SVM (5.65±5.12 dB) and MLR (6.96±5.38 dB) (all, p<0.001). Mean of point-wise mean AE with CNN was 5.47±3.05 dB, significantly smaller than those with SVM (7.96±4.63 dB) and MLR (11.71±4.15 dB) (all, p<0.001). R2 with CNN was 0.74 for the mean TH of whole VF, and 0.44±0.24 for the overall 68 points. CONCLUSION: DL model showed considerably accurate prediction of HFA 10-2 VF from SD-OCT.

来源:Br J Ophthalmol,查看原文。版权归原作者所有。

引用

  1. Hashimoto Y, Asaoka R, Kiwaki T, et al. Deep learning model to predict visual field in central 10° from optical coherence tomography measurement in glaucoma. Br J Ophthalmol. 2020. PMID: 32593978. doi:10.1136/bjophthalmol-2019-315600

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。