No abstract

No abstract

As a non-invasive angle closure glaucoma diagnosis procedure, Van Herick's slit-lamp Limbal Anterior Chamber Depth Estimation (LACDE) technique is preferably the current standard screening method preliminarily performed at hospitals and eye clinics. However, accuracy of depth estimation is very much dependent on the slit projection distance from the limbus (LD). This study investigated the relation and designed a parametric model to reduce the inconsistencies between results and improving the Van Herick's method accuracy. Additionally, two different types of classifiers such as Support Vector Machine (SVM) andDecision Tree with addition of LD as a predictor were trained and tested for comparison with parametric model. In the end acquired results clearly showed the effectiveness of LD consideration in LACDE, which lead to improvement in the experimental results.

No abstract

No abstract