Glistenings Detection Program

Glistenings Detection Program

The artificial intraocular lens (IOL) is inserted to replaced the crystalline lens of the human eye after cataract surgery or refraction clear lens extraction. When the IOL, is in an aquatic environment, glistenings, which are liquid-filled microvacuoles, can be discriminated from normal state. Automatic detection of glistenings is a new problem, they have tiny sizes, sometimes have low contrast and also similar with the lens background. In this paper, the candidate glistenings are automatically detected by mathematic morphology method and fine segmented using the classifiers, we used k-nearest neighbor (kNN) and Naïve Bayes for comparing the results. The detected glistenings are validated by object-based with ophthalmologists hand-drawn ground-truth. The result shows that classification can improve the performance of glistenings detection better than using only morphology method. The proposed software was developed for creating an effective automatic glistenings detection and quantification with a user-friendly Graphical User Interface, reliable results. Thus this computer-aid tool will help the researcher analyze the experiment results to better understand glistenings characteristic that have an effect on vary conditions.