Efficient Diabetic Retinopathy Space Subjective Fuzzy Clustering In Digital Fundus Images

Diabetic Retinopathy (DR) is a very rigorous and extensive eye disease which causes blindness in diabetic patients. DR primary cause for blindness is not for the reason that it has the highest occurrence but it often remains unobserved until severe vision loss occurs. Diabetic Retinopathy is characterized by changes in the retina that comprise of change in blood vessel diameter, micro aneurysms, lipid, and protein deposits with cotton coat spot. It depends on the appearance, hemorrhages and new vessel growth. For the further improvement in detection of DR, a new technique named Space Subjective Fuzzy Clustering Method (SSFC) is introduced which used for the extraction of retinal vessels. Initially the segmentation of digital fundus images takes place and then SSFC method introduced with mathematical study of shape. In mathematical study of shape, the digital fundus image is smoothed and strengthened so that the retinal vessels are enhanced and the backdrop information is suppressed. The Space Subjective Fuzzy Clustering algorithm is then employed to the enhanced segmented image. After the SSFC, a refinement process is used to decrease the fragile edges noise, and the final results of the retinal vessels of DR are accordingly achieved. The performance of the SSFC method is compared with some existing segmentation methods using Gold Standard Database, STARE, and DRIVE Dataset. Performance of Space Subjective Fuzzy Clustering Method is measured in term of segmentation accuracy ratio and clustering efficiency. The approach has been tested on a series of digital fundus images, and experimental results show that our technique is hopeful and effective.