Efficient and high accuracy 3-D OCT angiography motion correction in pathology

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Non-rigid motion artifacts in optical coherence tomography angiography (OCTA) volumes degrade image quality.
• Accurate 3D motion correction is crucial for quantitative analysis of retinal vasculature.
• Existing methods often struggle with aligning both axial and transverse features, especially in pathological cases.

Purpose of the Study:

• To develop and validate a novel, non-rigid 3D motion correction method for OCTA volumes.
• To achieve sub-pixel alignment and micrometer-scale distortion correction in all three dimensions.
• To create a robust and clinically applicable method for motion correction in OCTA.

Main Methods:

• A joint optimization approach aligning axial structural features (retinal layers) and transverse angiographic vascular features.
• Incorporation of orthogonal scanning and kinematically plausible displacement modeling.
• No segmentation of specific structures, ensuring robustness to pathological changes.

Main Results:

• Achieved sub-pixel alignment and micrometer-scale distortion correction in 3D.
• Demonstrated state-of-the-art axial motion correction.
• Showed significant improvements in transverse co-alignment and distortion correction, particularly in patients with pathologies.

Conclusions:

• The novel method effectively corrects non-rigid 3D motion in OCTA volumes.
• The approach is robust to pathologies and suitable for clinical integration due to its parallel implementation and short runtime.
• This advancement enhances the diagnostic potential of OCTA by improving image fidelity.