Mobile cataract detection using optimal combination of statistical texture analysis

Cataract is one of potentially dangerous disease that will be causing the blindness as an impact of the belated in handling cataract. Cataract is not only disrupting productivity and mobility of patients, but also causing the social-economic impact that will decrease the quality of life. Early detection of cataract reputed as a principal arrangement in restraining the increasing number of blindness caused by cataract. Commonly, an ophthalmologist uses a slit lamp camera to diagnose a cataract. Lacking of ophthalmologist and slit lamp camera in rural areas are the main problem of the belated in diagnosing cataract. In this paper, we investigate the optimal combination candidate of statistical texture features that is provide highest accuracy for cataract detection. In this research, we use K-Nearest Neighbor (k-NN) as classification method that will be implemented on android smartphone. Our result show that the optimal combination of texture features are dissimilarity, contrast, and uniformity. The highest accuracy of the system is 97.5%. The system is implemented on mobile smartphone.