OCT-GAN: Single Step Shadow and Noise Removal from Optical Coherence Tomography Images of the Human Optic Nerve Head
This addresses the need for faster and more accurate OCT image processing for clinicians and algorithms in ophthalmology, though it appears incremental as it builds on existing GAN-based methods for noise and shadow removal.
The paper tackles the problem of speckle noise and retinal shadows in OCT B-scans, which obscure important features and hinder diagnosis, by developing a single-step algorithm that removes both issues in 10.4ms. The result shows significant improvements: mean AGM increased by 57.2%, and PSNR, CNR, and SSIM increased by 11.1%, 154%, and 187% respectively compared to single-frame B-scans.
Speckle noise and retinal shadows within OCT B-scans occlude important edges, fine textures and deep tissues, preventing accurate and robust diagnosis by algorithms and clinicians. We developed a single process that successfully removed both noise and retinal shadows from unseen single-frame B-scans within 10.4ms. Mean average gradient magnitude (AGM) for the proposed algorithm was 57.2% higher than current state-of-the-art, while mean peak signal to noise ratio (PSNR), contrast to noise ratio (CNR), and structural similarity index metric (SSIM) increased by 11.1%, 154% and 187% respectively compared to single-frame B-scans. Mean intralayer contrast (ILC) improvement for the retinal nerve fiber layer (RNFL), photoreceptor layer (PR) and retinal pigment epithelium (RPE) layers decreased from 0.362 \pm 0.133 to 0.142 \pm 0.102, 0.449 \pm 0.116 to 0.0904 \pm 0.0769, 0.381 \pm 0.100 to 0.0590 \pm 0.0451 respectively. The proposed algorithm reduces the necessity for long image acquisition times, minimizes expensive hardware requirements and reduces motion artifacts in OCT images.