Software environment for CG noise creation.

PurposeRandom noise-moving images (noises) can make glaucoma patients with no subjective symptoms aware of visual field abnormalities. To explore this concept, we developed a noise using computer graphics (CG) and investigated the difference in the subjective perception of visual field abnormalities between CG and conventional analog noises.MethodsWe enrolled individuals with glaucoma (205 eyes), preperimetric glaucoma (PPG; 19 eyes), and normal eyes (35 eyes). For a CG noise, a series of still images was made by randomly selecting five monochromatic tones on 2-mm square dots, and these images were drawn at 60 frames per second (fps) to create a noise-moving image. The participants were asked to describe their perceived shadows on a paper. The results were categorized as follows based on the pattern deviation probability map of the Humphrey field analyzer (HFA): “agreement,” “partial agreement,” “disagreement,” and “no response.” The glaucoma stage was classified into four stages, from M1 to M4, based on the HFA’s mean deviation.ResultThe detection rates (agreement and partial agreement) were 80.5% and 65.4% for the CG and analog noises, respectively, with CG noise showing a significantly higher detection rate in all glaucoma eyes (P ConclusionThe CG noise is more effective than the analog noise in evaluating the subjective perception of visual field abnormalities in patients with glaucoma.