Development and validation of deep learning algorithms for automated eye laterality detection with anterior segment photography

Abstract

This paper aimed to develop and validate a deep learning (DL) model for automated detection of the laterality of the eye on anterior segment photographs. Anterior segment photographs for training a DL model were collected with the Scheimpflug anterior segment analyzer. We applied transfer learning and fine-tuning of pre-trained deep convolutional neural networks (InceptionV3, VGG16, MobileNetV2) to develop DL models for determining the eye laterality. Testing datasets, from Scheimpflug and slit-lamp digital camera photography, were employed to test the DL model, and the results were compared with a classification performed by human experts. The performance of the DL model was evaluated by accuracy, sensitivity, specificity, operating characteristic curves, and corresponding area under the curve values. A total of 14,468 photographs were collected for the development of DL models. After training for 100 epochs, the DL models of the InceptionV3 mode achieved the area under the receiver operating characteristic curve of 0.998 (with 95% CI 0.924–0.958) for detecting eye laterality. In the external testing dataset (76 primary gaze photographs taken by a digital camera), the DL model achieves an accuracy of 96.1% (95% CI 91.7%–100%), which is better than an accuracy of 72.3% (95% CI 62.2%–82.4%), 82.8% (95% CI 78.7%–86.9%) and 86.8% (95% CI 82.5%–91.1%) achieved by human graders. Our study demonstrated that this high-performing DL model can be used for automated labeling for the laterality of eyes. Our DL model is useful for managing a large volume of the anterior segment images with a slit-lamp camera in the clinical setting.

Automatic Classification of Slit-Lamp Photographs by Imaging Illumination

“…Li et al implemented a DL-based system to automatically detect poor quality slit-lamp images, which are commonly assessed manually in the clinical setting. 26,27 Zheng et al 28 developed a DL system that automates detection and annotation of eye laterality. Deshmukh et al 29 developed a DL algorithm that automatically segmented the corneal deposits in patients with corneal stromal dystrophy using slit-lamp images.…”

Automated image curation in diabetic retinopathy screening using deep learning

“…However, the laterality model also had excellent classification performance when stratified by retinal field (Supplementary Table S5 ). A reported DL model trained to classify laterality in anterior segment images alone achieved an AUROC of 0.998 41 . The classification of unidentifiable laterality images was excellent in the internal test set but reduced in the external dataset.…”

Development of a deep learning algorithm for myopic maculopathy classification based on OCT images using transfer learning

“…We used a t-distributed Stochastic Neighbor Embedding (t-SNE) method to reduce high-dimensional DL data features to two-dimensions and visualize them to observe the feature subspace aggregation capability of the algorithm. The t-SNE converts similarities between data points and minimizes the Kullback–Leibler divergence of the joint probabilities between the low-dimensional embedding and the high-dimensional data ( 19 ). The heat map analysis approach was consistent with our previous study ( 16 ).…”