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

Lin, Zhiwen; Guo, Ruoqian; Wang, Yanjie; Wu, Bian; Chen, Tingting; Wang, Wenzhe; Chen, Danny Z.; Wu, Jian

doi:10.1007/978-3-030-00934-2_9

Cited by 80 publications

(49 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This section verifies whether adding the synthesized images into training set for data augmentation can be beneficial for training DR grading models. We train 3 baseline models (VGG-19, ResNet-50 [ 53 ], Inception-v3), two competitive DR grading models AFN [ 54 ] and Zhou et al (DenseNet-121) [ 44 ] with and without data augmentation by synthesized images. We only modify the output dimension of the last fully connected layer to 5 for baseline models.…”

Section: Experiments and Resultsmentioning

confidence: 99%

RF‐GANs: A Method to Synthesize Retinal Fundus Images Based on Generative Adversarial Network

Chen

Long

Guo

2021

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

Diabetic retinopathy (DR) is a diabetic complication affecting the eyes, which is the main cause of blindness in young and middle-aged people. In order to speed up the diagnosis of DR, a mass of deep learning methods have been used for the detection of this disease, but they failed to attain excellent results due to unbalanced training data, i.e., the lack of DR fundus images. To address the problem of data imbalance, this paper proposes a method dubbed retinal fundus images generative adversarial networks (RF-GANs), which is based on generative adversarial network, to synthesize retinal fundus images. RF-GANs is composed of two generation models, RF-GAN1 and RF-GAN2. Firstly, RF-GAN1 is employed to translate retinal fundus images from source domain (the domain of semantic segmentation datasets) to target domain (the domain of EyePACS dataset connected to Kaggle (EyePACS)). Then, we train the semantic segmentation models with the translated images, and employ the trained models to extract the structural and lesion masks (hereafter, we refer to it as Masks) of EyePACS. Finally, we employ RF-GAN2 to synthesize retinal fundus images using the Masks and DR grading labels. This paper verifies the effectiveness of the method: RF-GAN1 can narrow down the domain gap between different datasets to improve the performance of the segmentation models. RF-GAN2 can synthesize realistic retinal fundus images. Adopting the synthesized images for data augmentation, the accuracy and quadratic weighted kappa of the state-of-the-art DR grading model on the testing set of EyePACS increase by 1.53% and 1.70%, respectively.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

RF‐GANs: A Method to Synthesize Retinal Fundus Images Based on Generative Adversarial Network

Chen

Long

Guo

2021

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

show abstract

“…Wang et al [9] incorporated attention mechanisms to state‐of‐the‐art deep models to focus the training on salient lesions. Last, Lin et al [10] augmente the annotation of the data set with bounding boxes of important lesions for some examples and use this annotation to enhance the DR detection. Adding microaneurysms segmentation embeddings to the classification embeddings improves the results for both screening and referable DR detection compared to the vanilla classification model.…”

Section: Methodsmentioning

confidence: 99%

“…This allows the model to make a more informed decision regarding DR grade. It was later proposed to use bounding boxes extra annotation for some lesions to enhance DR grading [10]. The development of reliable approaches for segmenting lesions responsible for the pathology is needed for more reliable and interpretable decision‐making diagnostic models.…”

Section: Introductionmentioning

confidence: 99%

Microaneurysms segmentation and diabetic retinopathy detection by learning discriminative representations

et al. 2020

View full text Add to dashboard Cite

Deep learning techniques are recently being used in fundus image analysis and diabetic retinopathy detection. Microaneurysms are important indicators of diabetic retinopathy progression. The authors introduce a two‐stage deep learning approach for microaneurysms segmentation using multiple scales of the input with selective sampling and embedding triplet loss. The proposed approach facilitates a region proposal fully convolutional neural network trained on segmented patches and a patch‐wise refinement network for improving the results suggested by the first stage hypothesis. To enhance the discriminative power of the second stage refinement network, the authors use triplet embedding loss with a selective sampling routine that dynamically assigns sampling probabilities to the oversampled class patches. This approach introduces a 23.5% relative improvement over the vanilla fully convolutional neural network on the Indian Diabetic Retinopathy Image Data set segmentation data set. The proposed segmentation is incorporated in a classification model to solve two downstream tasks for diabetic retinopathy detection and referable diabetic retinopathy detection. The classification tasks are trained on the Kaggle diabetic retinopathy challenge data set and evaluated on the Messidor data. The authors show that adding the segmentation enhances the classification performance and achieves comparable performance to the state‐of‐the‐art models.

show abstract

“…However, though achieving high sensitivity and specificity, these deep learning based methods lack the intuitive explanation for the decision. The recent methods [12]- [14] shifted the focus to locate the lesion position with a weakly supervised learning framework. However, these methods often relied on a large training set of lesion annotations from professional experts.…”

Section: A Diabetic Retinopathy Detectionmentioning

confidence: 99%

“…For the sake of generality, we select o O DR detector, a CNN based method within the top-3 entries on Kaggle's challenge. Even by now, the performance of o O is still competent to the latest method [14].…”

Section: A Diabetic Retinopathy Detectionmentioning

confidence: 99%

Explainable Diabetic Retinopathy Detection and Retinal Image Generation

Niu,

Gu,

Zhao

et al. 2021

Preprint

View full text Add to dashboard Cite

Though deep learning has shown successful performance in classifying the label and severity stage of certain diseases, most of them give few explanations on how to make predictions. Inspired by Koch's Postulates, the foundation in evidence-based medicine (EBM) to identify the pathogen, we propose to exploit the interpretability of deep learning application in medical diagnosis. By determining and isolating the neuron activation patterns on which diabetic retinopathy (DR) detector relies to make decisions, we demonstrate the direct relation between the isolated neuron activation and lesions for a pathological explanation. To be specific, we first define novel pathological descriptors using activated neurons of the DR detector to encode both spatial and appearance information of lesions. Then, to visualize the symptom encoded in the descriptor, we propose Patho-GAN, a new network to synthesize medically plausible retinal images. By manipulating these descriptors, we could even arbitrarily control the position, quantity, and categories of generated lesions.We also show that our synthesized images carry the symptoms directly related to diabetic retinopathy diagnosis. Our generated images are both qualitatively and quantitatively superior to the ones by previous methods. Besides, compared to existing methods that take hours to generate an image, our second level speed endows the potential to be an effective solution for data augmentation.Code is available at https://github.com/zzdyyy/ Patho-GAN.

show abstract

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

Cited by 80 publications

References 12 publications

RF‐GANs: A Method to Synthesize Retinal Fundus Images Based on Generative Adversarial Network

RF‐GANs: A Method to Synthesize Retinal Fundus Images Based on Generative Adversarial Network

Microaneurysms segmentation and diabetic retinopathy detection by learning discriminative representations

Explainable Diabetic Retinopathy Detection and Retinal Image Generation

Contact Info

Product

Resources

About