Generalized 3D registration algorithm for enhancing retinal optical coherence tomography images.

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Optical coherence tomography (OCT) is crucial for diagnosing and monitoring ocular disorders.
• Motion artifacts degrade OCT image quality, limiting clinical utility.
• Existing registration algorithms are often system-specific, hindering broad application.

Purpose of the Study:

• To develop a system-agnostic B-scan registration algorithm for motion artifact removal in OCT.
• To create artifact-free OCT volumes independent of specific system parameters or acquisition protocols.

Main Methods:

• A B-scan registration algorithm was developed to correct translational and rotational eye movements.
• The algorithm was tested on diverse datasets from custom and commercial OCT systems with varying specifications.
• Accuracy was quantified using Mean Squared Error (MSE) and Multiscale Structural Similarity Index Measure (MS-SSIM).

Main Results:

• The algorithm significantly improved image contrast, enabling detailed visualization of superficial and deep retinal vasculature.
• Finer retinal features, including photoreceptors and pathologies, became discernible after registration and averaging.
• Quantitative analysis showed decreased MSE and increased MS-SSIM with more averaged registered volumes.

Conclusions:

• The developed algorithm provides superior visualization of retinal microvasculature and morphology.
• The system-agnostic approach enhances the clinical utility of OCT across different modalities.
• This methodology improves OCT's role in diagnosing and monitoring ocular pathologies.