Purpose:
To compare the performance of goniophotography and EyeCam for angle closure diagnosis, using gonioscopy as the reference standard.
Methods:
In this prospective observational study, consecutive subjects recruited from a glaucoma clinic underwent gonioscopy by a single glaucoma specialist. EyeCam (Clarity Medical System, Pleasanton, CA) and goniophotography (Topcon Corporation, Tokyo, Japan) were performed in all quadrants of one randomly selected eye and the images were evaluated in a random sequence on different days by a single observer, masked to clinical findings. Angle closure was defined as an eye in which the pigmented trabecular meshwork was not visible for ≥180 degrees- this definition was used for all 3 methods. Agreement between methods was analyzed by kappa statistics and area-under receiver operating characteristic curves (AUC) were obtained, with gonioscopy as the reference standard.
Results:
Of the 85 subjects recruited, there were 38 eyes with angle closure by gonioscopy. The majority of subjects were Chinese (74.1%) and female (52.9%). The agreement for diagnosis of angle closure between gonioscopy and goniophotography (k=0.86) and between gonioscopy with EyeCam (k=0.86) were very good; AUC was also similar for both methods (AUC = 0.93 for both methods, p=1.0). However, the agreement for angle closure diagnosis between goniophotography and EyeCam was less (k = 0.72; 95% CI, 0.57 - 0.87). Overall, there were more eyes diagnosed as angle closure using either imaging method compared to gonioscopy but this was not significant (p=0.69, Figure).
Conclusions:
Goniophotography and EyeCam showed good agreement with gonioscopy for angle closure diagnosis.
Keywords: imaging/image analysis: clinical • anterior segment • clinical (human) or epidemiologic studies: systems/equipment/techniques