Automatic tracking of pupillary dynamics from in vivo functional optical coherence tomography images

Abstract

A major challenge currently faced in measuring visually evoked optical changes in different retinal layers using in vivo and non-invasive functional optical coherence tomography (fOCT) is the need for pupillary dilation drugs to enable reliable measurement of the retinal functional responses. However, there are situations where the use of pupillary dilation drugs is not suitable in both clinical and animal ophthalmic research situations. To overcome this challenge, we propose a new approach to tracking visually evoked pupillary dynamics, based on the use of a fast scanning fOCT system and a novel computerised framework, which can then be used to isolate the pure stimulus-evoked functional response from the retina imaged using the same system. The proposed framework tracks the highly reflective edges of the iris in the cross-sectional fOCT tomograms by applying a three-step strategy: (a) a noise-compensation reconstruction in order to reduce the effect of speckle noise on the pupil edge tracking accuracy; (b) a region-based maximum likelihood tracking of the pupil edge and (c) an outlier compensation for calculation of the pupil size.