Comparison of structure-function correlation among IMO visual function analyser and Humphrey field analyser.

Journal: The British Journal Of Ophthalmology
Published:
Abstract

Objective: To compare the structure-function correlations between optical coherence tomography (OCT) thickness parameters and the corresponding sectorial mean sensitivity (MS) values obtained with IMO visual function analyser (IMOvifa) and Humphrey field analyser (HFA) perimeters.

Methods: Prospective cross-sectional study. Methods: Glaucoma suspect, glaucoma and ocular hypertension patients underwent IMOvifa 24-2 Ambient Interactive Zippy Estimated Sequential Testing, HFA 24-2 Swedish Interactive Threshold Algorithm-Standard and OCT tests within 12 weeks. Based on the Garway-Heath map, the sensitivity points were divided in corresponding OCT peripapillary retinal nerve fibre layer (pRNFL) sectors, and also the central 10° points in corresponding ganglion cell-inner plexiform layer thickness over the macula. The structure-function correlations were analysed by Spearman's rank coefficient and compared by Steiger's test.

Results: 57 eyes of 57 patients (mean age 59.2 years, 61.4% female) with reliable field and adequate OCT strength were included in this analysis. IMOvifa had significantly shorter testing time (3.7 vs 5.3 mins), lower mean deviation (difference -0.6 dB) and greater visual field index (difference +1.0%) compared with HFA (p<0.001 for all). There was moderate to strong correlation between IMOvifa and HFA MS sectors (p<0.001 for all). IMOvifa had significant correlations in the inferior (I, r=0.46), superior (S, r=0.45), temporal (T, r=0.28) pRNFL sectors and in the I (r=0.63) and S (r=0.53) mGCPIPL hemispheres (p≤0.03 for all). There was no significant difference in the strength of the correlation of IMOvifa versus HFA results with OCT parameters (p≥0.06).

Conclusions: The IMOvifa test took less time and resulted in slightly less severe defects than HFA. MS in all sectors was moderately to strongly correlated. Function-structure correlations were similar when using either perimeter.

Authors
Joyce Kang, Rafaella Nascimento E Silva, Julia Kim, Dolly Chang, Yihao Li, Chen Chen, Yan Zhao, Sofia De Arrigunaga, Sandra Freeman, Tobias Elze, Michael Lin, David Friedman