False-positive patterns in optical coherence tomography in hyperopic and myopic eyes.

Abstract

This study aimed to evaluate how axial length influences false-positive rates in ganglion cell analysis (GCA) and retinal nerve fibre layer (RNFL) maps and to identify distinct false-positive patterns on optical coherence tomography (OCT) deviation maps across axial length subgroups. A retrospective cross-sectional study of 194 healthy eyes included spectral-domain OCT imaging. False positives were defined as colour-coded areas on GCA/RNFL maps. Eyes were stratified into three axial length groups: group A (<23 mm), group B (23 mm≤axial length<25 mm) and group C (≥25 mm). Abnormal OCT deviation map patterns were analysed and compared across axial length groups. False-positive results were observed in 76.1% of cases for GCA and 62.2% for RNFL maps. Rates increased significantly with axial length: for GCA deviation maps, the rates were 45.2% in group A, 51.2% in group B and 87.5% in group C (p<0.001). The most frequent false-positive pattern in hyperopic eyes was the small doughnut type (67.9%), while diffuse thinning and inferior crescent patterns predominated in highly myopic eyes (both 38.1%). In logistic regression analysis, axial length was an independent predictor of false positives in all maps (OR 1.09-2.64, p<0.05). Older age also correlated with false positives (OR 1.04-1.06, p<0.05), except for minimum ganglion cell-inner plexiform layer thickness. Abnormal OCT diagnostic classification should be interpreted with caution, particularly in eyes with long axial length and in older patients. Distinctive patterns in the GCA deviation map are observable in patients with markedly short or long axial lengths, and recognising these false-positive patterns may prompt more cautious interpretation.