A model to estimate the change in photoreceptor outer segment length that occurs upon light stimulation – a biomarker of cone function in the retina.

Vision relies on the isomerization of photopigments within photoreceptors (cones and rods) by light. Traditionally, researchers and clinicians have relied on functional assessments of photoreceptor function such as electroretinography (ERG) to diagnose and treat retinal disease. While these techniques are sensitive to photoreceptor loss, their resolution is insufficient to discern functional changes in individual photoreceptors. Retinal imaging with adaptive optics (AO)-enhanced ophthalmoscopes is noninvasive and has the resolution necessary to observe human photoreceptors at a cellular scale and detect structural abnormalities; however, there is a need for methods that analyze this imaging data as a noninvasive alternative to quantifying visual function.

A model-based analytical tool for AO data that estimates the change in photoreceptor outer segment length upon light stimulation rather than standardizing reflectance to a pre-stimulation interval. The photoreceptor change in length is associated with cellular function, so this estimation provides a non-invasive biomarker of cone function with increased precision, signal to noise ratio, customizability, and high spatial resolution.

A method for extracting the functional responses of single cone photoreceptors to light stimulation from AO imaging data. Previous methods estimate single cone responses by quantifying variability in the amount of light reflected at each time point. The method described here provides a model-based approach of analyzing the same measurements to describe the effect of a change in cone optical path length (OPL) on the reflected light, accounting for the heterogeneity of the response across acquisitions. By fitting the model to the data from multiple acquisitions for each cone, that change in cone OPL over time is estimated directly.

• This method demonstrates an improved signal-to-noise ratio compared to the existing variability-based method.
• Functional results obtained via this model may be compared against and/or calibrated with measurements of the same phenomena using other measurement modalities.
• The physically-based analysis approach may be more readily associated with the underlying biophysical mechanisms that mediate cone function
• Single-cone AO optoretinography is noninvasive and offers the possibility of functional measurements in the areas of the human retina with the highest cone density and smallest cone size.
• This model can be customized by changing key parameters, optimizing the basic formulation to different conditions, generalizing across various metrics, or varying the data fitting procedure, among others.