Automated Vessel Segmentation of Retinal Vessels using Hybrid Region Information

Kumar Parasuraman, B. Selva Gayathri, K. Senthamarai Kannan



Recent researches state that the categorization of retinal vessels into artery/vein (A/V) is a critical phase for automating the detection of vascular adjustments and for the calculation of function symptoms related to numerous systemic illnesses which includes diabetes, high blood pressure, and different cardiovascular situations. This paper affords an automatic method for A/V class, primarily based on the evaluation of a graph extracted from the retinal vasculature. The proposed method classifies the entire vascular tree choosing the kind of each intersection factor (graph nodes) and assigning one of two labels to every vessel phase (graph links). Very last category of a vessel phase as A/V is achieved through the combination of the graph primarily based labeling results with a fixed of intensity features. The proposed model applies it to three open retinal picture datasets (two datasets of shading fundus photography and one fluorescein angiography dataset). The proposed model beats its rivals when analyzed with other generally utilized unsupervised and administered techniques. For test, the affectability (0:742), particularity (0:982) and precision (0:954) achieved on the DRIVE dataset are close to second onlooker's explanations. File terms vessel, division, neighborhood stage.

 Keywords: A/V: Artery and Vein, optical disc, ANN, GMM, GRF

Cite this Article

Kumar Parasuraman, Selva Gayathri B, Senthamarai Kannan K. Automated Vessel Segmentation of Retinal Vessels using Hybrid Region Information. Research & Reviews: Journal of Statistics. 2018; 7(3): 28–36p.

Full Text:




  • There are currently no refbacks.