Novel Approaches to Improve Microaneurysm Detection in Retinal Images

In this paper, we present a novel approach to improve microaneurysm candidate extraction in color fundus images. The individual algorithms published so far can be hardly considered in an automatic screening system. To improve further the sensitivity, specificity and image classification rate of microaneurysm detection we propose an appropriate combination of individual algorithms. Thus, we investigate the detection of microaneurysms through the following phases: first, we use different approaches to extract microaneurysm candidates. Then, we select candidates voted by a sufficient number of candidate extractor algorithms. Finally, we classify the candidates with a machine-learning based approach. Our framework improves the positive likelihood ratio and outperforms both state-of-the-art individual candidate extractors and microaneurysm detectors in these terms.