Quantitative assessment of retinal attenuation and backscattering in OCT imaging using iterative layer-based analysis

Optical coherence tomography (OCT) generates cross-sectional images by measuring backscattered light from within tissue, with image contrast influenced by tissue optical properties, primarily optical scattering and illumination conditions. The optical attenuation coefficient (OAC), derived from the...

Full description

Saved in:
Bibliographic Details
Published inBiomedical optics express Vol. 16; no. 7; pp. 2808 - 2823
Main Authors Zhang, Yi, Gong, Zhaoyu, Shi, Yaping, Le, Viet Hoan, Pan, Hefu, Liu, Jian, Wang, Ruikang K.
Format Journal Article
LanguageEnglish
Published United States Optica Publishing Group 01.07.2025
Online AccessGet full text

Cover

Loading…
More Information
Summary:Optical coherence tomography (OCT) generates cross-sectional images by measuring backscattered light from within tissue, with image contrast influenced by tissue optical properties, primarily optical scattering and illumination conditions. The optical attenuation coefficient (OAC), derived from the OCT signal, can decouple illumination effects and directly probe the scattering properties, offering potential functional contrast. However, the existing OAC quantification methods are prone to bias in multilayered structures, such as retinal tissue, primarily due to the inhomogeneity in backscattering fraction. In this study, we introduce a layer-based iterative method that improves OAC accuracy by accounting for layer-specific backscattering fractions. Our approach refines OAC calculations through an optimized depth-resolved framework, iteratively adjusting the residual light term for improved estimation. We validated the method’s convergence and accuracy through both mathematical derivation and numerical simulations. Further, we evaluated its clinical applicability in healthy subjects. Comparative results demonstrated that our method enhances tissue characterization, highlighting its potential to advance retinal imaging and disease assessment.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
These authors contribute equally to this paper
ISSN:2156-7085
2156-7085
DOI:10.1364/BOE.563121