Purpose:
To find out the characteristics of visual field abnormality of Frequency Doubling Technology (FDT) perimetry in glaucoma and non-glaucoma eyes in a population based study so as to find out a more effective evaluation way using FDT perimetry in glaucoma screening.
Methods:
Population-based cross-sectional study. The study population consisted of 4439 subjects aged 40+ years in rural and urban areas of Beijing. All participants underwent an ophthalmological examination including fundus photography and FDT perimetry. Glaucoma was defined by a glaucomatous optic disc appearance. Reliable results of FDT were defined as false positive rate and fixation loss not more than 1/3. Percentage of abnormality and its correlation with glaucoma in any location of the 17 FDT testing grids were analyzed separately. Hierarchical clustering and discriminate analysis were applied.
Results:
4350 (98.0%) subjects (8617 eyes) provided measurement data with readable photographs and reliable FDT results, in which 905 eyes showed abnormal FDT results and 205 eyes were diagnosed as glaucoma. The positive percentage of each testing location ranged from 1.8% to 5.1%. According to our numbering rule, grid 1, 2, 4, 5, 12, 13 had positive judging power and grid 0, 11 had negative in glaucoma prediction. The 3 grids in nasal inferior were strongly related with glaucoma. After using the algorithm to detect glaucoma in population, the falsely classification percentage was 4.5%.
Conclusions:
Each testing location of FDT had ranging contributions to glaucoma predicting. Nasal inferior grids showed a more convincing value in glaucoma screening, while the central 2 locations had negative predictive value. We conducted a discriminate algorithm whose accuracy to predict glaucoma in populations increased from 90.2% to 95.5%, which showed a more promising judging power.
Keywords: perimetry • visual fields