Automated image processing pipeline for adaptive optics scanning light ophthalmoscopy.

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Adaptive optics scanning light ophthalmoscopy (AOSLO) is a powerful imaging technique for retinal visualization.
• The clinical utility of AOSLO is currently limited by a substantial post-processing workload.
• Existing processing methods are often manual, time-consuming, and require specialized expertise.

Purpose of the Study:

• To develop and validate an open-source, automated image processing pipeline for AOSLO.
• To reduce the post-processing burden and improve the efficiency of AOSLO data analysis.
• To enhance the clinical accessibility and reduce the operating costs of AOSLO imaging.

Main Methods:

• Development of an automated AOSLO image processing pipeline with "live" and "full" processing modes.
• The "live" mode provides real-time feedback during image acquisition.
• The "full" mode integrates disparate processing modules for automated montage generation.

Main Results:

• The automated pipeline significantly reduced human operator time by 54.9% ± 28.4%.
• No significant differences were observed in cone density metrics compared to manual processing.
• The mean lag between initiation and montage placement in live mode was 54.6s ± 32.7s.

Conclusions:

• The developed automated AOSLO processing pipeline effectively mitigates the post-processing burden.
• This automation decreases technical overhead and operating costs, enhancing clinical accessibility.
• The open-source pipeline offers a scalable solution for routine AOSLO data analysis.