Le Christopher, Baroni Mariana, Vinnett Alfred, Levin Moran R, Martinez Camilo, Jaafar Mohamad, Madigan William P, Alexander Janet L
Department of Ophthalmology and Visual Sciences, University of Maryland School of Medicine, Baltimore, MD, USA.
Department of Ophthalmology, Children's National Medical System, Washington, DC, USA.
Transl Vis Sci Technol. 2020 Dec 23;9(2):63. doi: 10.1167/tvst.9.2.63. eCollection 2020 Dec.
Ultrasound biomicroscopy (UBM) is a noninvasive method for assessing anterior segment anatomy. Previous studies were prone to intergrader variability, lacked assessment of the lens-iris diaphragm, and excluded pediatric subjects. Lens status classification is an objective task applicable in pediatric and adult populations. We developed and validated a neural network to classify lens status from UBM images.
Two hundred eighty-five UBM images were collected in the Pediatric Anterior Segment Imaging Innovation Study (PASIIS) from 80 eyes of 51 pediatric and adult subjects (median age = 4.6 years, range = 3 weeks to 90 years) with lens status phakic, aphakic, or pseudophakic ( = 33, 7, and 21 subjects, respectively). Following transfer learning, a pretrained Densenet-121 model was fine-tuned on these images. Metrics were calculated for testing dataset results aggregated from fivefold cross-validation. For each fold, 20% of total subjects were partitioned for testing and the remaining subjects were used for training and validation (80:20 split).
Our neural network trained across 60 epochs achieved recall 96.15%, precision 96.14%, F1-score 96.14%, false positive rate 3.74%, and area under the curve (AUC) 0.992. Feature saliency heatmaps consistently involved the lens. Algorithm performance was compared using 2 image sets, 1 from subjects of all ages, and the second from only subjects under age 10 years, with similar performance under both circumstances.
A neural network trained on a relatively small UBM image set classified lens status with satisfactory recall and precision. Adult and pediatric image sets offered roughly equivalent performance. Future studies will explore automated UBM image classification for complex anterior segment pathology.
Deep learning models can evaluate lens status from UBM images in adult and pediatric subjects using a limited image set.
超声生物显微镜检查（UBM）是一种用于评估眼前节解剖结构的非侵入性方法。以往的研究容易出现分级者差异，缺乏对晶状体-虹膜隔的评估，并且排除了儿科受试者。晶状体状态分类是一项适用于儿科和成人人群的客观任务。我们开发并验证了一种神经网络，用于从UBM图像中对晶状体状态进行分类。
在儿科眼前节成像创新研究（PASIIS）中，收集了51名儿科和成人受试者（年龄中位数 = 4.6岁，范围 = 3周龄至90岁）的80只眼睛的285张UBM图像，晶状体状态为有晶状体眼、无晶状体眼或人工晶状体眼（分别为33、7和21名受试者）。在迁移学习之后，对一个预训练的Densenet-121模型在这些图像上进行微调。计算从五折交叉验证汇总得到的测试数据集结果的指标。对于每一折，将20%的总受试者划分用于测试，其余受试者用于训练和验证（80:20分割）。
我们的神经网络在60个轮次上进行训练，召回率达到96.15%，精确率达到96.14%，F1分数达到96.14%，假阳性率为3.74%，曲线下面积（AUC）为0.992。特征显著性热图始终涉及晶状体。使用2个图像集比较算法性能，1个来自所有年龄段的受试者，另1个仅来自10岁以下的受试者，在两种情况下性能相似。
在相对较小的UBM图像集上训练的神经网络对晶状体状态进行分类时，召回率和精确率令人满意。成人和儿科图像集的性能大致相当。未来的研究将探索针对复杂眼前节病变的UBM图像自动分类。
深度学习模型可以使用有限的图像集从成人和儿科受试者的UBM图像中评估晶状体状态。
