Closed-loop optical stabilization and digital image registration in adaptive optics scanning light ophthalmoscopy

Area of Science:

• Ophthalmology
• Biomedical Engineering
• Optical Imaging

Background:

• Acquiring high-resolution retinal images using adaptive optics scanning light ophthalmoscopy (AOSLO) is significantly challenged by involuntary eye motion.
• Existing methods struggle to fully compensate for the dynamic nature of eye movements during imaging.
• Precise retinal imaging is crucial for diagnosing and monitoring various ocular diseases.

Purpose of the Study:

• To develop and implement a novel system for stabilizing retinal images acquired with AOSLO.
• To overcome the limitations imposed by eye motion in high-resolution retinal imaging.
• To achieve sub-micrometer accuracy in corrected retinal image registration.

Main Methods:

• Replaced the conventional scanning mirror with a two-axis tip/tilt mirror for simultaneous scanning and optical stabilization.
• Implemented a closed-loop optical stabilization system to actively counteract eye movements.
• Integrated real-time digital image registration to correct residual motion post-optical stabilization.

Main Results:

• Closed-loop optical stabilization reduced eye-movement-induced image motion amplitude by a factor of 10-15.
• Residual motion after optical stabilization was approximately 1.66-2.56 μm (0.34-0.53 arcmin).
• The combined optical and digital stabilization achieved an unprecedented accuracy of ~0.20-0.25 μm (0.04-0.05 arcmin) RMS.

Conclusions:

• The integrated system of optical stabilization and digital image registration effectively overcomes eye motion artifacts in AOSLO.
• This approach significantly enhances the quality and accuracy of high-resolution retinal imaging.
• The achieved accuracy represents a substantial advancement in retinal imaging technology, surpassing previous methods.