A Comparative Study of Methods for Fundus Mask Generation

  • Mr. Manohar P, Dr. Mallikarjun S Holi

Abstract

Digital fundus imaging is a non-invasive imaging modality wherein a fundus camera provides a magnified and detailed view of the eye fundus that is broadly utilized for large scale screening of ocular diseases for example, Diabetic Retinopathy (DR). The DR is a vision threatening, tiny blood vessel entanglement of diabetes wherein the back of the eye called retina is damaged as a consequence of liquids oozing from circulatory paths into the retina. Prevention and effective treatment of sight threatening DR requires early diagnosis. The study region of the retinal ocular fundus is the Field of View (FOV) that includes anatomic features and lesion abnormalities which must be processed by the detection and feature extraction algorithms. The dark area surrounding the effective FOV must be discarded. This necessitates the generation of an appropriate fundus mask. The latest research still struggle with the issue of fundus mask generation mainly due to the dearth of comparative studies, reviews of literature and normative datasets. The paper focuses on the implementation and comparative study of state of art methodologies which are grouped into three categories based on the threshold value selection for binary fundus mask generation. All methods are assessed on a publicly available DRIVE database.

 

Keywords: Comparative study, Diabetic Retinopathy, Fundus mask, Preprocessing, Retinal image analysis, Thresholding.

Published
2020-06-06
How to Cite
Mr. Manohar P, Dr. Mallikarjun S Holi. (2020). A Comparative Study of Methods for Fundus Mask Generation. International Journal of Advanced Science and Technology, 29(3), 9642 - 9653. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/26934
Section
Articles