Sub2Full: split spectrum to boost OCT despeckling without clean data

TL;DR

Sub2Full introduces a self-supervised OCT despeckling method that leverages spectrum splitting of repeated scans to improve image quality without needing clean training data.

Contribution

It proposes a novel spectrum-splitting strategy for self-supervised OCT denoising, enabling high-quality despeckling without clean datasets.

Findings

Outperforms Noise2Noise and Noise2Void in visual quality

Effective in visualizing retinal sublaminar structures

Validated on vis-OCT retinal images

Abstract

Optical coherence tomography (OCT) suffers from speckle noise, causing the deterioration of image quality, especially in high-resolution modalities like visible light OCT (vis-OCT). The potential of conventional supervised deep learning denoising methods is limited by the difficulty of obtaining clean data. Here, we proposed an innovative self-supervised strategy called Sub2Full (S2F) for OCT despeckling without clean data. This approach works by acquiring two repeated B-scans, splitting the spectrum of the first repeat as a low-resolution input, and utilizing the full spectrum of the second repeat as the high-resolution target. The proposed method was validated on vis-OCT retinal images visualizing sublaminar structures in outer retina and demonstrated superior performance over conventional Noise2Noise and Noise2Void schemes. The code is available at https://github.com/PittOCT/Sub2Full-OCT-Denoising.