A Vessel Keypoint Detector for Junction Classification
Published in 14th IEEE International Symposium on Biomedical Imaging (ISBI 2017), 2017
Recommended citation: C. L. Srinidhi, P. Rath, J. Sivaswamy (2017). "A Vessel Keypoint Detector for Junction Classification", 2017 IEEE 14th International Symposium on Biomedical Imaging (ISBI 2017)
The human eye is an incredibly complex object, and many changes that occur in the retina can be linked to various physical diseases. This paper investigates a new method for detecting vessel junctions in the human retina, which are linked to said diseases. Using a new feature vector called the Vessel Keypoint Detector, the system first identifies vessel junction points using a combination of Computer Vision and Machine Learning techniques. After identification, the system also classifies the vessel junctions as either bifurcations or crossovers, since both are important for tracking different classes of diseases. Additionally, the system may be deployed for performing vessel analysis for biometric tracking.
You can view the paper here.