A robust and reliable quantification method for Focal Arteriolar Narrowing in color retinal image

In this paper a reliable and robust method is presented for the quantification of Focal Arteriolar Narrowing (FAN), a precursor for hypertension, stroke and other cardiovascular diseases. Our contribution in this paper is that we have proposed a novel edge based retinal blood vessel segmentation technique which is very effective in low contrast retinal images. In addition to that we developed a robust and reliable measurement technique to quantify Focal Arteriolar Narrowing. For initial results and quantitative evaluation of the proposed method, we evaluate our proposed method on a dataset of 53 manually graded vessel segments. The experimental results indicate a strong correlation between the computed Focal Arteriolar Narrowing (FAN) values and the expert grading (Spearman coefficient of 0.76, P <; 0.0001). In addition, the results also show that the system can detect healthy and FAN affected cases with an accuracy of 93%. This has demonstrated the reliability of the proposed method for automatic focal narrowing assessment. The quantitative measurements provided by the system may help to establish a more reliable link between Focal Arteriolar Narrowing (FAN) and known systemic and eye diseases, which will be investigated further.