Fundus Image Quality Assessment: a Brief Review of Techniques

Image quality assessment (IQA) of the inner surface of the eye opposite the lens (includes the retina as well as the optic disc, macula, fovea, and posterior pole) is an important component of training artificial neural networks used in clinical decision support systems for diagnosing ophthalmic diseases. IQA helps to determine the quality of the incoming image, which can then be either included in the training sample or enhanced to achieve correct neural network scores and reduce the probability of diagnostic error. In the brief review the most frequently used IQA techniques are presented, their weaknesses and strengths are clarified, and the prospects for the development of IQA for retinal images are given.