Adaptive classifier allows enhanced flow contrast in OCT angiography using a histogram-based motion threshold and 3D...

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Ophthalmology

Background:

• Optical coherence tomography angiography (OCTA) is crucial for visualizing retinal vasculature.
• Accurate flow contrast enhancement in OCTA is challenging due to depth-dependent artifacts and complex vessel structures.

Purpose of the Study:

• To propose an adaptive digital classifier for superior flow contrast enhancement in OCTA.
• To address depth dependence and vessel shape variations in OCTA image analysis.

Main Methods:

• A depth-adaptive motion threshold was determined using histogram analysis and fitting to distinguish static and dynamic voxels.
• A modified vesselness function and 3D Hessian analysis were employed for shape-based classification to remove residual static voxels.
• The classifier integrates both motion and 3D shape information for robust flow detection.

Main Results:

• The proposed adaptive classifier effectively overcomes depth-dependent limitations in motion-based OCTA analysis.
• Shape-based classification successfully adapts to deformed vessel shapes, improving accuracy.
• Experimental results demonstrate superior flow contrast enhancement compared to existing methods.

Conclusions:

• The adaptive digital classifier significantly enhances flow contrast in OCTA by leveraging motion and 3D shape characteristics.
• This method offers a promising tool for more accurate and reliable OCTA image interpretation.
• Improved flow contrast aids in the diagnosis and monitoring of various ocular diseases.