Speckle reduction in optical coherence tomography by image registration and matrix completion

Area of Science:

• Biomedical Imaging
• Ophthalmology
• Medical Image Processing

Background:

• Speckle noise significantly degrades image quality in Optical Coherence Tomography (OCT).
• Swept source OCT (SS-OCT) uses multiple scans for speckle reduction, but eye movement complicates this process.

Purpose of the Study:

• To develop an advanced method for speckle reduction in multiply-scanned OCT images.
• To overcome challenges posed by eye movement during rapid retinal scanning.

Main Methods:

• A novel speckle reduction technique involving preliminary reduction, global and local image registration (fast iterative diamond search), and low rank matrix completion with bilateral random projection.
• Iterative estimation of noise and recovery of the underlying clean image.

Main Results:

• The proposed method achieved an average contrast-to-noise ratio of 15.65.
• This represents a significant improvement over the baseline method's 13.78 in swept source OCT devices.

Conclusions:

• The developed technology effectively reduces speckle noise in OCT images.
• This method can be integrated into existing OCT machines to enhance image quality for improved diagnostic analysis.