Three-Dimensional Retinal Microvasculature Restoration in OCT Angiography

Researchers have developed a deep learning algorithm that restores three-dimensional capillary anatomy from a single optical coherence tomographic angiography (OCTA) volume, addressing imaging artifacts that hinder reliable quantification of retinal blood flow. The model, using an EfficientNet-B5 encoder and squeeze-and-excitation modules, achieved significant improvements in image quality and microvascular fidelity, with peak signal-to-noise ratio rising from 22.23 to 26.16 and structural similarity index from 0.72 to 0.91. The technique also increased Dice coefficient overlap by at least 3.8% in 2D and 51.2% in 3D across multiple vascular slabs, offering a method to restore intrinsic three-dimensional retinal microvasculature from standard OCTA scans.

arXiv:2606.05375v1 Announce Type: new Abstract: Optical coherence tomographic angiography OCTA is a powerful technique for imaging retinal microvasculature. However, acquiring reliable quantification of retinal blood flow and areas of retinal nonperfusion is challenging because of imaging artifacts. Existing methods primarily focus on noise suppression, projection artifact removal, or signal enhancement to improve the image quality of OCTA in cross-sectional or two-dimensional 2D en face projections, while neglecting the intrinsic three-dimensional vascular architecture. In this study, we propose a deep learning-based algorithm for restoring capillary anatomical vasculature from a single OCTA volume. The network consists of an EfficientNet-B5 encoder and a decoder incorporating concurrent spatial and channel squeeze-and-excitation modules, connected via skip connections to preserve spatial resolution. Three adjacent B-frames are used as input to predict the restored middle B-frame. We evaluated the performance of the model using the peak signal-to-noise ratio PSNR and structural similarity index measure SSIM against ground truth generated from averaging multiple scans. The results show that the proposed model significantly both p < 0.001 improved image quality compared with the original single OCTA volume, with a PSNR of 26.16 +/- 1.26 vs. 22.23 +/- 0.78 and an SSIM of 0.91 +/- 0.02 vs. 0.72 +/- 0.03. The proposed model also significantly p < 0.001 improved microvascular fidelity, measured by the Dice coefficient overlap between the model output and ground truth, in both 2D and 3D by at least 3.8% and 51.2%, respectively, across several different vascular slabs.