Abstract:Accurate multi-lesion segmentation together with automated grading on fundus images played a vital role in diagnosing and treating diabetic retinopathy (DR). Nevertheless, the intrinsic patterns of fundus lesions aggravated challenges in DR detection process. Therefore, we proposed a novel multi-lesion segmentation guided deep attention network (MSGDA-Net) for accurate and automated DR detection, consisting of a DR lesion segmentation pathway as an auxiliary task to produce a lesion regional prior knowledge and a DR grading pathway to extract local fine-grained features and long-range dependency. In DR lesion segmentation pathway, we designed a Multi-Scale Attention Block (MSAB) and a Lesion-Aware Relation Block (LARB) to allow interactions among multi-lesion features for alleviating ambiguity in lesion segmentation, generating lesion regional prior knowledge. As for DR grading pathway, we presented a Spatial-Fusion Block (SFB) to enhance the lesion-related local fine-grained feature representations and eliminate irrelevant noise information under the guidance of the resulting lesion regional priors, while constructed an Enhanced Self-Attention Block (ESAB) to optimally fuse fine-grained features from SFB with long-range global-context information for grading DR. The experimental results showed that our MSGDA-Net not only achieved state-of-the-art performance in the tasks of multi-lesion segmentation and DR grading, reaching up to 49.21 % Dice, 38.05 % IoU and 51.15 % AUPR for DR lesion segmentation on the DDR dataset, as well as accuracy values of 75.00 % and 87.18 % for DR grading on local newly-built VisionDR and publicly available APTOS datasets, but also manifested good generalization and robustness on cross-data evaluation. It could serve as a promising tool for computer aided DR screening and diagnosis.

Mask Region-oriented Diabetic Retinopathies Detection in Ophthalmic Medical Images Via Non-local Attention.

Lesion-aware attention network for diabetic nephropathy diagnosis with optical coherence tomography images

CLC-Net: Contextual and Local Collaborative Network for Lesion Segmentation in Diabetic Retinopathy Images

Pseudo-Labeling for Small Lesion Detection on Diabetic Retinopathy Images

Multi-lesion segmentation guided deep attention network for automated detection of diabetic retinopathy

Lesion-aware network for diabetic retinopathy diagnosis

Zoom-in-Net: Deep Mining Lesions for Diabetic Retinopathy Detection

An attentional mechanism model for segmenting multiple lesion regions in the diabetic retina

Mini Lesions Detection on Diabetic Retinopathy Images via Large Scale CNN Features

Diabetic Retinopathy Severity Identification Using 3D Dual-Domain Attention Approach

A Multi-Scale Self-Attention Network for Diabetic Retinopathy Retrieval.

Diabetic Retinopathy Detection Based on Deep Convolutional Neural Networks for Localization of Discriminative Regions

DRDr: Automatic Masking of Exudates and Microaneurysms Caused By Diabetic Retinopathy Using Mask R-CNN and Transfer Learning

A Deep Convolutional Neural Network for Diabetic Retinopathy Detection Via Mining Local and Long‐range Dependence

Effective Methods of Diabetic Retinopathy Detection Based on Deep Convolutional Neural Networks

Classification of diabetic retinopathy algorithm based on a novel dual-path multi-module model

A Novel Lesion Detection Algorithm Based on Multi-Scale Input Convolutional Neural Network Model for Diabetic Retinopathy

A Framework for Identifying Diabetic Retinopathy Based on Anti-noise Detection and Attention-Based Fusion.

Diabetic Retinopathy Detection Based on Weakly Supervised Object Localization and Knowledge Driven Attribute Mining.

HACDR-Net: Heterogeneous-Aware Convolutional Network for Diabetic Retinopathy Multi-Lesion Segmentation

Automatic Segmentation of Diabetic Macular Edemas and ILM-RPE Layers in Retinal OCT Images Based on MFANet.