Automated quality evaluation of digital fundus photographs

Abstract

. Purpose: Retinal images acquired by means of digital photography are often used for evaluation and documentation of the ocular fundus, especially in patients with diabetes, glaucoma or age‐related macular degeneration. The clinical usefulness of an image is highly dependent on its quality. We set out to develop and evaluate an automatic method of evaluating the quality of digital fundus photographs. Methods: A method for making a numerical quantification of image sharpness and illumination was developed using Matlab™ image analysis functions. Based on their sharpness and illumination measures, 1000 fundus photographs, randomly selected from a clinical database, were assigned to four predefined quality groups (not acceptable, acceptable, good, very good). Six independent observers, comprising three experienced ophthalmologists and three ophthalmic nurses with extensive experience in fundus image acquisition, classified a selection of 100 of these images into the corresponding quality groups. Results: Automatic quality evaluation was more sensitive than evaluation by human observers in terms of ability to discriminate between good and very good images. The median concordance between the six human observers and the automatic evaluation was substantial (kappa = 0.64). Conclusions: The proposed method provides an objective quality assessment of digital fundus photographs which agrees well with evaluations made by qualified human observers and which may be useful in clinical practice.

Automated Brightness and Contrast Adjustment of Color Fundus Photographs for the Grading of Age-Related Macular Degeneration

“…With so many pitfalls, substandard images have been reported to be as high as 20% in clinical studies. 11,14 Even with best practice, acquisition of consistently high-quality images cannot be guaranteed at the time of capture. Post hoc standardization 15 is therefore necessary for AMD grading.…”

Retinal image restoration by means of blind deconvolution

“…Image quality is subjectively evaluated by the person capturing the images, and they can sometimes mistakenly accept a low-quality image. 3 Low-quality image occurrence rate has been reported at 3.7-19.7% in clinical studies, 4-6 which is not a minor fact. A recent study by Abràmoff et al 7 using an automated system for detection of diabetic retinopathy found that from 10,000 exams 23% had insufficient image quality.…”

“…8 Moreover, regardless of how well controlled the aforementioned parameters are, in practice it may not always be possible to obtain good enough image quality as a result of additional factors such as lens opacities in the examined eye, scattering, insufficient pupil dilation or patient difficulty in steady fixating a target in the camera (such as in patients suffering from amblyopia). 3 Out of all possible retinal image degradations, some can be properly compensated via enhancement or restoration techniques (e.g., low-contrast, nonuniform illumination, noise, and blur). 2 However, this compensation is also dependent on the extent of the degradation.…”

Quality evaluation of digital fundus images through combined measures

“…[6][7][8][9][10] A shortcoming of local analysis is the processing time that, usually, is longer than in global techniques. Segmentation of retinal features such as optic disc, fovea, and retinal vasculature is also included in some methods to augment specificity of the algorithms to fundus images.…”