Purchase this article with an account.
J. F. Updike, P. G. Updike, S. R. Sadda, A. C. Walsh; Quantifying Fixation Instability During Fourier Domain OCT. Invest. Ophthalmol. Vis. Sci. 2007;48(13):159. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Eye movements during OCT data acquisition may interfere withaccurate retinal thickness map reconstructions and intervisitcomparisons of thickness measurements. In this study, eye movementsduring 3D Fourier Domain OCT (3D-OCT) capture were analyzedand rectifed by aligning fundus video images that had been acquiredconcurrently with the OCT data.
Seventeen infrared fundus video sequences from 25 eyes of 14consecutive patients imaged with a prototype 3D-OCT device (Topcon,Japan) were aligned using custom software that measures subpixeloffsets between frames. The intended point of fixation for eachcase was estimated by the average position of the fundus duringthe sequence. Offsets from this point of fixation were analyzedover time and compared to logMAR visual acuity.
The average visual acuity in this cohort of patients was 20/80.The average deviation from the intended point of fixation was136 microns. The actual fundus position was more than 100 micronsfrom intended point of fixation for 50% of the 3D-OCT scan (Figure1). Fixation instability occurred throughout the scan and didnot correlate with acquisition time. Horizontal eye movementscorrelated well (r=0.69, p<.01) with logMAR visual acuity(Figure 2).
Substantial eye movements can be expected during 3D-OCT acquisition.Along with other factors, these movements may depend on subjectvisual acuity but not acquisition time.
This PDF is available to Subscribers Only