Abstract
Purpose: :
A method for classifying cone photoreceptors has been demonstrated, relying on the spectral properties of the cone classes and the absorptive effects of bleaching [1]. This method, however, requires multiple images over 24 hours to achieve sufficient signal–to–noise. Recently, we have shown that several unique patterns of cone scintillation, near 1 Hz, emerge under certain imaging conditions [2]. An interference model, based on refractive index changes of the outer segment during bleaching, predicts similar unique scintillation signatures for the three cone classes [3]. This model may allow rapid cone classification if techniques are developed to correlate observed and predicted scintillations. Such techniques are evaluated here using experimental data.
Methods: :
Cone mosaic images were acquired using an adaptive optics retina camera (described in [2] and [3]) at 30 Hz over 3 sec intervals, with a 670 nm source used for simultaneous bleaching and imaging. Registered images representing the temporal mean and RMS intensity level were generated. Cone identification was automated, and temporal reflectance profiles were extracted. Characteristics of the profiles were computed in order to partition the cones according to scintillation temporal dynamics.
Results: :
Distributions of scintillation characteristics were noticeably clustered for those profiles with high fluctuations, e.g. the 20% of the cones with highest signal. Clustering reduced as the number of cones considered was increased.
Conclusions: :
Methods for classifying cones based upon bleaching–dependent scintillation show promise as a way of classifying cone photoreceptors using a 3 second measurement. Moreover these methods allow direct measurement of the ratio of cones of different classes, assuming that the magnitude of the signal generated by an individual cone is uncorrelated with its class. [1] A. Roorda and D. R. Williams, Nature 397, 520–522 (1999).
[2] J. Rha et al., presented at Frontiers in Optics, 88th OSA Annual Meeting, Rochester, NY (2004).
[3] J. Rha et al., Invest. Ophthalmol. Vis. Sci. 46, E–Abstract 3546 (2005).
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • photoreceptors • image processing