Automated Detection of Microaneurysms in Color Fundus Images using Deep Learning with Different Preprocessing Approaches

TL;DR

This study compares two preprocessing techniques, Illumination Equalization and Top-hat transformation, combined with deep learning for automated microaneurysm detection in retinal images, achieving around 80-90% accuracy compared to manual detection.

Contribution

It introduces a combined approach of preprocessing, vessel segmentation, and deep learning for microaneurysm detection, evaluating the impact of different preprocessing methods.

Findings

Illumination equalization preprocessing yields about 90% detection accuracy.

Top-hat transformation preprocessing achieves over 80% accuracy.

The method outperforms manual detection in large retinal image datasets.

Abstract

Imaging methods by using computer techniques provide doctors assistance at any time and relieve their workload, especially for iterative processes like identifying objects of interest such as lesions and anatomical structures from the image. Detection of microaneurysms (MAs) as one of the lesions in the retina is considered to be a crucial step in some retinal image analysis algorithms for the identification of diabetic retinopathy (DR) as the second-largest eye diseases in developed countries. The objective of this study is to compare the effect of two preprocessing methods, Illumination Equalization, and Top-hat transformation, on retinal images to detect MAs using a combination of Matching based approach and deep learning methods either in the normal fundus images or in the presence of DR. The steps for the detection are as following: 1) applying preprocessing, 2) vessel segmentation and masking, and 3) MAs detection using a combination of Matching based approach and deep learning. From the accuracy viewpoint, we compared the method to manual detection performed by ophthalmologists for our big retinal image databases (more than 2200 images). Using first preprocessing method, Illumination equalization and contrast enhancement, the accuracy of MAs detection was about 90% for all databases (one local and two publicly retinal databases). The performance of the MAs detection methods using top-hat preprocessing (the second preprocessing method) was more than 80% for all databases.