[MS03] Biomedical Imaging in Disease Diagnosis

[MS03-03] Modified B-COSFIRE filters for segmentation of the retinal blood vessels

*Varun Makkar¹, Rajesh Kumar Pandey¹, B. V. Rathish Kumar² (1. Indian Institute of Technology (BHU) Varanasi (India), 2. Indian Institute of Technology Kanpur (India))

Keywords:

Retinal blood vessel segmentation,B-COSFIRE filters,Centreline detection

Retinal blood vessel segmentation is crucial for detecting and monitoring eye diseases such as diabetic retinopathy, glaucoma, and hypertensive retinopathy. It helps in analyzing retinal vascular structures, identifying abnormalities, and tracking disease progression. Automated segmentation enhances diagnostic accuracy, reduces manual effort, and supports telemedicine applications. In this work, we propose a novel algorithm for retinal blood vessel segmentation. The algorithm starts by enhancing the retinal vasculature by using non-isotropic diffusion based B-COSFIRE filters. Final vessel segments are obtained by applying hysteresis thresholding to the combined response of the B-COSFIRE filters and centreline algorithm on the input fundus image. The proposed algorithm achieves state-of-the-art performance in the class of unsupervised segmentation algorithms. It is tested on publically available datasets like DRIVE and STARE, using multiple performance metrics.