Detection of neovascularisation using K-means clustering through registration of peripapillary OCT and fundus retinal images

This paper focuses on quantitatively assessing the presence of neovascularization by registration of peripapillary Optical Coherence Tomography (OCT) and fundus images of Diabetic Retinopathy patients. Here in this present work, the two intra-patient images acquired by spectral domain OCT modality and fundus camera are enhanced and the blood vessels are extracted using kirsch template. Then the two images are fused by similarity measure based registration. The automatic algorithm for extraction of neovascularization features using k-means clustering is proposed to quantify and detect normal and abnormal blood vessels. Results shows that the proposed method produces accurate results for all the input real time sample images and the results are validated with experts clinical findings.