Reproduce “Low-rank MERA Decomposition of 3D OCT Image Volume for Compression, Denoising, Feature Extraction and Diagnosis of Retinal Diseases” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ivica Kopriva, Fei Shi, Marija Štanfel, Qin Peng, Jianhao Bao, Gaowei Li, Xinjian Chen
MERA compresses retinal OCT volumes with a low-rank, multi-scale tensor representation that models correlations across and within B-scans, suppresses some speckle, and provides features for classification. Its volume representation is training-free, although classification uses labeled examples. On Topcon scans, segmentation error was 1.87% of retinal thickness at compression ratio 10, showing preserved segmentation in this setting—not universal benefit.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.