A Deep Learning–Based Fully Automated Program for Choroidal Structure Analysis Within the Region of Interest in Myopic Children

Abstract

Purpose To develop and validate a fully automated program for choroidal structure analysis within a 1500-µm-wide region of interest centered on the fovea (deep learning–based choroidal structure assessment program [DCAP]). Methods A total of 2162 fovea-centered radial swept-source optical coherence tomography (SS-OCT) B-scans from 162 myopic children with cycloplegic spherical equivalent refraction ranging from −1.00 to −5.00 diopters were collected to develop the DCAP. Medical Transformer network and Small Attention U-Net were used to automatically segment the choroid boundaries and the nulla (the deepest point within the fovea). Automatic denoising based on choroidal vessel luminance and binarization were applied to isolate choroidal luminal/stromal areas. To further compare the DCAP with the traditional handcrafted method, the luminal/stromal areas and choroidal vascularity index (CVI) values for 20 OCT images were measured by three graders and the DCAP separately. Intraclass correlation coefficients (ICCs) and limits of agreement were used for agreement analysis. Results The mean ± SD pixel-wise distances from the predicted choroidal inner, outer boundary, and nulla to the ground truth were 1.40 ± 1.23, 5.40 ± 2.24, and 1.92 ± 1.13 pixels, respectively. The mean times required for choroidal structure analysis were 1.00, 438.00 ± 75.88, 393.25 ± 78.77, and 410.10 ± 56.03 seconds per image for the DCAP and three graders, respectively. Agreement between the automatic and manual area measurements was excellent (ICCs > 0.900) but poor for the CVI (0.627; 95% confidence interval, 0.279–0.832). Additionally, the DCAP demonstrated better intersession repeatability. Conclusions The DCAP is faster than manual methods. Also, it was able to reduce the intra-/intergrader and intersession variations to a small extent. Translational Relevance The DCAP could aid in choroidal structure assessment.

Longitudinal Changes in Choroidal Structure Following Repeated Low-Level Red-Light Therapy for Myopia Control: Secondary Analysis of a Randomized Controlled Trial

“…Details of the program has been described elsewhere in our previous publication. 20 A validation study has shown that the automatic measurements of LA, SA, TCA, and CVI are highly correlated with manual measurements and have reduced intraobserver, interobserver, and intersession variations to a small extent. 20…”

“…18,19 We developed a fully automated and efficient deep learningbased choroidal structure assessment program with proven accuracy in comparison with manual measurement tool. 20 Repeated low-level red-light (RLRL) therapy has been confirmed as an effective and safe intervention for myopia control in children. [21][22][23][24][25][26][27] In addition, a clinically significant subfoveal choroidal thickening has been noted following RLRL therapy, [21][22][23][24]28 with the magnitude of the thickening associated with RLRL therapy's efficacy in myopia control.…”

Multi-Dataset Comparison of Vision Transformers and Convolutional Neural Networks for Detecting Glaucomatous Optic Neuropathy from Fundus Photographs

“…ViT models have now been applied to the analysis and interpretation of a wide range of clinical data ranging from electrocardiograms [28] to intraoperative surgical techniques [29]. In ophthalmology, there are increasing reports of ViT models trained to classify retinal pathologies from fundus photography [30][31][32] and OCT imaging [33][34][35][36], including several assessing their performance relative to CNNs [31,34,35]. Given that glaucoma diagnosis often requires multimodal imaging that correlates structural and functional data, it has been theorized that the global attention mechanisms utilized by ViTs offer an advantage over CNNs' dependency upon local features.…”