The Impact of Torsional Head Movement on Disc to Fovea Angle Measurements Used to Correct Optical Coherence Tomography Scans

Abstract

 

Purpose

 

The between-subject variation in disc-to-fovea angle (DFA) during OCT scans is due to torsional head movement as well as true anatomical differences. To determine the relative contribution of each, fundus images obtained during repeat fdOCT scans were analyzed.

 

Methods

 

Fundus images for 13 glaucoma patients were selected from a large database based on availability of at least 2 high-quality repeat images from at least 2 visits. Data from a single visit for 15 healthy controls were also obtained. The center of the fovea, identified on fundus photos by using co-registered b-scans from macular cube scans (3DOCT-1000, Topcon, NJ), and the center of the disc were marked on the fundus photos using a Matlab program and the DFA determined (Fig. 1). For each image, 3 measurements were made; the reproducibility was good (mean SD of 0.13°). The average of the 3 measurements was taken as the DFA of that image. The standard deviation (SD) and coefficient of variation (CV) were used as descriptive statistics of variability. For each eye, the within-subject variability was defined as the SD and CV of the DFA across visits. The between-subject variability was defined as the SD and CV of the mean DFA for all patients, averaged across repeat images from all visits for a given patient. A Monte Carlo simulation (n=100,000) was used to determine statistical significance.

 

Results

 

The average DFA for controls (7.90°, range 0.58° to 13.97°) was similar to the DFA for glaucoma patients (7.50°, range -0.45° to 14.56°). The between-subject variability (SD=4.10°, CV=55%) was significantly (p≈0.005) larger than the within-subject variability (SD=1.25°, CV=17%). The absolute within-subject difference in DFA between visits ranged from 0.22° to 4.46°.

 

Conclusions

 

The within-subject variability of DFA was significantly smaller than the between-subject variability, suggesting that variation in DFA across scans is due largely to true anatomical differences. However, for some individual patients, head tilt can play a major role. Because scan rotation does not correct fully for anatomical differences [1], the best approach to improving OCT scans may be to develop procedures for minimizing head tilt during scanning, rather than using tilted-scan protocols. 1. Hood et al. PRER 2012.

 

 

Fig. 1: Fundus view showing disc-to-fovea angle (DFA).

 

Keywords: 552 imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)