Fully automated identification and clinical classification of macular edema using optical coherence tomography images

Abstract Diabetic macular edema is a relevant ocular disease associated with diabetes mellitus that constitutes a concerning global health issue. This disease is one of the main causes of reversible blindness in industrialized countries, despite the availability of effective health interventions. This macular disorder is a consequence of the appearance of abnormal fluid regions within the main tissues of the retina that significantly decrease the patient's visual acuity. These fluid accumulations are also known as macular edemas (MEs). In this chapter, we present a computational methodology for the identification and clinical classification of ME using optical coherence tomography (OCT) scans, following the clinical classification of reference in the ophthalmological field. The presented system was tested with a dataset composed of 170 OCT scans retrieved from different patients that were labeled by a clinical expert. The presented system obtained satisfactory results in the localization of the area affected by each type of ME, even when they are combined in the same region of the retina. This fully automatic tool has demonstrated to be very useful not only for patients (helping in the early diagnosis of diseases that, consequently, improve their quality of life and well-being), but also in clinical systems by reducing costs.