A Deep Learning-Based Unified Framework for Red Lesions Detection on Retinal Fundus Images (2109.05021v6)

Abstract: Red-lesions, microaneurysms (MAs) and hemorrhages (HMs), are the early signs of diabetic retinopathy (DR). The automatic detection of MAs and HMs on retinal fundus images is a challenging task. Most of the existing methods detect either only MAs or only HMs because of the difference in their texture, sizes, and morphology. Though some methods detect both MAs and HMs, they suffer from the curse of dimensionality of shape and colors features and fail to detect all shape variations of HMs such as flame-shaped. Leveraging the progress in deep learning, we proposed a two-stream red lesions detection system dealing simultaneously with small and large red lesions. For this system, we introduced a new ROIs candidates generation method for large red lesions on fundus images; it is based on blood vessel segmentation and morphological operations, and reduces the computational complexity, and enhances the detection accuracy by generating a small number of potential candidates. For detection, we proposed a framework with two streams. We used pretrained VGGNet as a backbone model and carried out several extensive experiments to tune it for vessels segmentation and candidates generation, and finally learning the appropriate mapping, which yields better detection of the red lesions comparing with the state-of-the-art methods. The experimental results validated the effectiveness of the system in the detection of both MAs and HMs; it yields higher performance for per lesion detection; its sensitivity equals 0.8589 and good FROC score under 8 FPIs on DiaretDB1-MA reports FROC=0.7518, and with SN=0.7552 and good FROC score under 2,4and 8 FPIs on DiaretDB1-HM, and SN=0.8157 on e-ophtha with overall FROC=0.4537 and on ROCh dataset with FROC=0.3461 which is higher than the state-of-the art methods. For DR screening, the system performs well with good AUC on DiaretDB1-MA, DiaretDB1-HM, and e-ophtha datasets.