Spatial-frequency-based image reconstruction to improve image contrast in multi-offset adaptive optics ophthalmoscopy

Area of Science:

• Ophthalmology
• Optical Engineering
• Neuroscience

Background:

• Adaptive optics (AO) ophthalmoscopy enhances visualization of retinal structures.
• Off-axis detection methods improve contrast of translucent retinal cells.

Purpose of the Study:

• To develop a 2D optical model of off-axis detection in AO ophthalmoscopy.
• To investigate the impact of offset orientation on phase contrast.
• To introduce a post-processing method for improved cellular contrast.

Main Methods:

• Developed a 2D optical model for phase contrast in off-axis AO ophthalmoscopy.
• Acquired in vivo human retinal images using a multi-offset AO scanning light ophthalmoscope.
• Implemented a spatial-frequency-based image reconstruction post-processing technique.

Main Results:

• Phase contrast is dependent on the off-axis detector orientation.
• One axis yields high contrast due to asymmetric light distribution; the perpendicular axis yields low contrast due to symmetric distribution.
• The proposed reconstruction method significantly increased structural cellular contrast in human retinal neurons.

Conclusions:

• Off-axis detection orientation critically influences phase contrast in AO ophthalmoscopy.
• Spatial-frequency-based image reconstruction effectively combines multi-offset data.
• This approach substantially enhances in vivo imaging of retinal neurons, including cone inner segments, putative rods, and retinal ganglion cells.