Sparse Representation Classifier for microaneurysm detection and retinal blood vessel extraction

Diabetic retinopathy (DR) is a common complication of diabetes that damages the retina and leads to sight loss if treated late. In its earliest stage DR can be diagnosed by the presence of a microaneurysm (MA). Although some algorithms have been developed thus far, the accurate detection of MA in a color retinal image is still a challenging problem. By locating a blood vessels' width, color, reflectivity, tortuosity, and abnormal branching, one can deduce the existence of DR. In this paper we propose new methods to detect MA based on Dictionary Learning (DL) with Sparse Representation Classifier (SRC), and extract retinal blood vessels using SRC. For MA detection we first locate all possible MA candidates with Multi-scale Gaussian Correlation Filtering (MSCF), and then classify these candidates with Dictionary Learning (DL) via SRC. Particularly, two dictionaries one for the MA and one for the non-MA are learned from example MA and non-MA structures, and are used in the SRC process. Experimental results on the ROC database show that the proposed method can well distinguish MA from non-MA objects. Vessel extraction is based on Multi-scale Production of Matched Filter (MPMF) and SRC. First, we locate vessel center-line candidates using Multi-scale Matched Filtering, scale production, double thresholding and center-line detection. Then, the candidates which are center-line pixels are classified using SRC. Two dictionary elements of vessel and non-vessel are used in the SRC process. Experimental results on two public databases show that the proposed method is good at distinguishing vessel from non-vessel objects, and is even able to extract the center-line of small vessels.