Automated detection of age-related macular degeneration in color fundus photography: a systematic review

Abstract

The rising prevalence of age-related eye diseases, particularly age-related macular degeneration, places an ever-increasing burden on health care providers. As new treatments emerge, it is necessary to develop methods for reliably assessing patients' disease status and stratifying risk of progression. The presence of drusen in the retina represents a key early feature in which size, number, and morphology are thought to correlate significantly with the risk of progression to sight-threatening age-related macular degeneration. Manual labeling of drusen on color fundus photographs by a human is labor intensive and is where automatic computerized detection would appreciably aid patient care. We review and evaluate current artificial intelligence methods and developments for the automated detection of drusen in the context of age-related macular degeneration.

Keywords: age-related disorders; age-related macular degeneration; artificial intelligence; deep learning; machine learning.

Fig. 1

Illustration of supervised machine learning pipeline. 1) Image preprocessing is performed to reduce noise and enhance image features. 2) Features such as measures of entropy, energy, color and texture of image intensities, and spatial or geometric properties are extracted. 3) Features are grouped as numerical vectors (forming the image representation) and often undergo a selection process to decide which features best represent the image. 4) Training phase builds a model that tries to separate the data into the target, distinct classes. 5) The classifier—the mathematical function—that implements classification and defines the classes. 6) Testing is performed by classifying unseen data belonging to know classes.

Fig. 2

An overview of the ML methods in discussion and where they are applied at each stage. Deep Convolutional Neural Networks is a DL technique. ARMD, age-related macular degeneration; DL, deep learning; HSV, hue, saturation, value; ML, machine learning; RGB, red, green, blue; SVM, support-vector machine.