System and method for automated detection of age related macular degeneration and other retinal abnormalities

A method for automated detection of retinal abnormalities, comprising:

• receiving retinal data of the retina of a subject from a retinal scanner, wherein the retinal scanner is in communication with a computer;
• forming, by the computer, a retinal image based on the retinal data;
• locating, by the computer, one or more edges at each level of a pyramid decomposition of the retinal image;
• obtaining, by the computer, one or more binary mask images where the one or more edges are located;
• calculating, by the computer, a final mask as a logical union of the one or more binary mask images;
• creating, by the computer, a testing mask by morphologically closing the final mask;
• creating, by the computer, a training mask by taking a logical complement of a dilated final mask;
• and
• determining an area of abnormality in the retinal image by training an anomaly detector on the training mask and applying the anomaly detector on the testing mask using the computer.

A non-transitory computer-readable medium storing executable instructions for execution by a computer having memory, the medium storing instructions for to:

• receive retinal data from a retinal scanner coupled to the computer;
• form a retinal image based on the retinal data;
• locate one or more edges at each level of a pyramid decomposition of the retinal image;
• obtain one or more binary mask images where the one or more edges are located;
• calculate a final mask as a logical union of the one or more binary mask images;
• create a testing mask by morphologically closing the final mask;
• create a training mask by taking a logical complement of a dilated final mask;
• and
• determine an area of abnormality in the retinal image by training an anomaly detector on the training mask and applying the anomaly detector on the testing mask.