A Comprehensive Texture Segmentation Framework for Segmentation of Capillary Non-Perfusion Regions in Fundus Fluorescein Angiograms

Abstract

Capillary non-perfusion (CNP) in the retina is a characteristic feature used in the management of a wide range of retinal diseases. There is no well-established computation tool for assessing the extent of CNP. We propose a novel texture segmentation framework to address this problem. This framework comprises three major steps: pre-processing, unsupervised total variation texture segmentation, and supervised segmentation. It employs a state-of-the-art multiphase total variation texture segmentation model which is enhanced by new kernel based region terms. The model can be applied to texture and intensity-based multiphase problems. A supervised segmentation step allows the framework to take expert knowledge into account, an AdaBoost classifier with weighted cost coefficient is chosen to tackle imbalanced data classification problems. To demonstrate its effectiveness, we applied this framework to 48 images from malarial retinopathy and 10 images from ischemic diabetic maculopathy. The performance of segmentation is satisfactory when compared to a reference standard of manual delineations: accuracy, sensitivity and specificity are 89.0%, 73.0%, and 90.8% respectively for the malarial retinopathy dataset and 80.8%, 70.6%, and 82.1% respectively for the diabetic maculopathy dataset. In terms of region-wise analysis, this method achieved an accuracy of 76.3% (45 out of 59 regions) for the malarial retinopathy dataset and 73.9% (17 out of 26 regions) for the diabetic maculopathy dataset. This comprehensive segmentation framework can quantify capillary non-perfusion in retinopathy from two distinct etiologies, and has the potential to be adopted for wider applications.

Optical coherence tomography angiography versus fluorescein angiography in the diagnosis of ischaemic diabetic maculopathy

“…For example, Zheng et al. () proposed a new texture segmentation framework for the analysis of capillary non‐perfusion regions in FA which was satisfactory when compared to a reference standard of manual delineations. In contrast, OCT angiography automatically calculates the ischaemic areas in superficial and deep images.…”

Texture Image Segmentation Using Affinity Propagation and Spectral Clustering

“…(2) Construct the a±nity matrix A 2 R nÂn where its ith row and jth column element is de¯ned by Eq. (31).…”

“…So the texture image segmentation aims at segmenting a textured image into several regions each of which has the same or similar texture feature, where texture is very important information in image processing since it presents the interest of most objects and often indicates contextual information of image pixels. 31 Because segmentation of texture images depends on pixel patterns that are di±cult to capture or de¯ne, the techniques using pixel patterns for texture images are more di±cult than the techniques of only using the intensity of pixels. There have been many di®erent approaches for segmentation, 3,12,[18][19][20]30,31 and some belong to supervised methods, while others belong to unsupervised methods.…”

An automated framework to quantify areas of regional ischemia in retinal vascular diseases with OCT angiography

“…Fundus fluorescein angiography (FA) is considered to be the gold standard for imaging patients with retinal pathologies . Few studies have shown automated segmentation of CNP regions in fundus fluorescein angiography (FFA) images .…”