Automatic Fundus Image Segmentation for Diabetic Retinopathy Diagnosis by Multiple Modified U-Nets and SegNets

Ananda, Swathi; Kitahara, Daichi; Hirabayashi, Akira; Reddy, K. Rasool

doi:10.1109/apsipaasc47483.2019.9023290

Cited by 8 publications

(5 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the testing phase, we have used different datasets, to ensure that the result of the hemorrhages assessment is not biased due to the change of the dataset in the training and testing phases. Many similar studies on CAD system for hemorrhage diagnosis has recently been published [30,34,35,37] these studies allow us make a comparison in order to assess the performance of the proposed approch.…”

Section: Resultsmentioning

confidence: 99%

“…The sensitivity obtained was lower than the sensitivity of the proposed system with a value of 48.83. Ananda et al [37] suggested a methodology based on CNN by segmentation. They used the IDRiD and MESSIDOR databases, as well as a modified U-Net and a modified SegNet.…”

Section: Resultsmentioning

confidence: 99%

“…The network, called "SegNet", includes an encoder in the form of a 13-layer convolutional VGGNet and a decoder that manages the classification on a pixel-by-pixel basis. Ananda et al [37] modified U-Net deep neural network by reducing the number of filters and coding layers to segment different types of retinal diseases; each of the U-Net models is used to segment one of the disease types, such as hemorrhages, microaneurysms, hard exudates or soft exudates and optic disc. The modified version of the U-Net model yielded a dice coefficient of 0.86 for bleeding segmentation.…”

Section: Related Previous Workmentioning

confidence: 99%

See 2 more Smart Citations

Hemorrhage semantic segmentation in fundus images for the diagnosis of diabetic retinopathy by using a convolutional neural network

et al. 2022

View full text Add to dashboard Cite

Because retinal hemorrhage is one of the earliest symptoms of diabetic retinopathy, its accurate identification is essential for early diagnosis. One of the major obstacles ophthalmologists face in making a quick and effective diagnosis is viewing too many images to manually identify lesions of different shapes and sizes. To this end, researchers are working to develop an automated method for screening for diabetic retinopathy. This paper presents a modified CNN UNet architecture for identifying retinal hemorrhages in fundus images. Using the graphics processing unit (GPU) and the IDRiD dataset, the proposed UNet was trained to segment and detect potential areas that may harbor retinal hemorrhages. The experiment was also tested using the IDRiD and DIARETDB1 datasets, both freely available on the Internet. We applied preprocessing to improve the image quality and increase the data, which play an important role in defining the complex features involved in the segmentation task. A significant improvement was then observed in the learning neural network that was able to effectively segment the bleeding and achieve sensitivity, specificity and accuracy of 80.49%, 99.68%, and 98.68%, respectively. The experimental results also yielded an IoU of 76.61% and a Dice value of 86.51%, showing that the predictions obtained by the network are effective and can significantly reduce the efforts of ophthalmologists. The results revealed a significant increase in the diagnostic performance of one of the most important retinal disorders caused by diabetes.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Related Previous Workmentioning

confidence: 99%

See 1 more Smart Citation

Hemorrhage semantic segmentation in fundus images for the diagnosis of diabetic retinopathy by using a convolutional neural network

et al. 2022

View full text Add to dashboard Cite

show abstract

“…A novel cross‐disease attention network is developed in Reference 25 for screening Diabetic Macular Edema (DME) and DR, which has the capability of learning features of both diseases simultaneously. Ananda et al 26 proposed the multiple modified U‐Net and segmentation network for multi‐class segmentation tasks by allocating individual CNN to each type of disease.…”

Section: Related Workmentioning

confidence: 99%

A deeply supervised maximum response texton basedSegNetfor simultaneous multi retinal lesion segmentation

Pappu¹,

Krishna²,

Biswal³

et al. 2022

Int J Imaging Syst Tech

View full text Add to dashboard Cite

Diabetic Retinopathy (DR) is a diabetic mellitus complication that causes vision impairment and may lead to permanent blindness. The early signs of DR that appear on the retinal surface are microaneurysms, hemorrhages, hard exudates, and soft exudates. Hence the automatic detection of these retinal lesions assists in the early diagnosis of DR. This paper presents a novel deep learning model, MRT‐SegNet (Maximum Response Texton – Segmentation Network) for the automatic segmentation of different retinal lesions simultaneously along with the optic disc. In the proposed MRT‐SegNet, each encoder block consists of an MRT filter bank that extracts the textural feature maps of the retinal images and then fuses them with the local feature maps that are extracted from the traditional encoder block of the network. This fusion enables the network to segment the minute lesions from the retinal surface. The proposed model is evaluated on the IDRiD dataset and achieves a mean Area Under the Precision & Recall Curve (mAUC_PR) of 0.698 and AUC_PR scores of 0.495, 0.706, 0.823, 0.769 for microaneurysms, hemorrhages, hard exudates, and soft exudates respectively. The experimental results demonstrate that the MRT‐SegNet outperformed other multi retinal lesion segmentation models by achieving superior performance.

show abstract

“…To increase the feature map resolution, the original SegNet used an encoder to obtain the feature maps and employed a decoder to up-sample the feature maps ( 26 ). SegNet was first proposed by Saha et al ( 27 ) for road and indoor scene segmentation, and Ananda et al ( 28 ) introduced SegNet for DR image segmentation. To make optimal use of the global feature in image segmentation tasks, a global pyramid pooling layer and certain new strategies were proposed in PSPNet and compared with FCN ( 29 ).…”

Section: Introductionmentioning

confidence: 99%

Neighbored-attention U-net (NAU-net) for diabetic retinopathy image segmentation

Zhao,

Guan,

et al. 2023

Front. Med.

View full text Add to dashboard Cite

BackgroundDiabetic retinopathy-related (DR-related) diseases are posing an increasing threat to eye health as the number of patients with diabetes mellitus that are young increases significantly. The automatic diagnosis of DR-related diseases has benefited from the rapid development of image semantic segmentation and other deep learning technology.MethodsInspired by the architecture of U-Net family, a neighbored attention U-Net (NAU-Net) is designed to balance the identification performance and computational cost for DR fundus image segmentation. In the new network, only the neighboring high- and low-dimensional feature maps of the encoder and decoder are fused by using four attention gates. With the help of this improvement, the common target features in the high-dimensional feature maps of encoder are enhanced, and they are also fused with the low-dimensional feature map of decoder. Moreover, this network fuses only neighboring layers and does not include the inner layers commonly used in U-Net++. Consequently, the proposed network incurs a better identification performance with a lower computational cost.ResultsThe experimental results of three open datasets of DR fundus images, including DRIVE, HRF, and CHASEDB, indicate that the NAU-Net outperforms FCN, SegNet, attention U-Net, and U-Net++ in terms of Dice score, IoU, accuracy, and precision, while its computation cost is between attention U-Net and U-Net++.ConclusionThe proposed NAU-Net exhibits better performance at a relatively low computational cost and provides an efficient novel approach for DR fundus image segmentation and a new automatic tool for DR-related eye disease diagnosis.

show abstract

Automatic Fundus Image Segmentation for Diabetic Retinopathy Diagnosis by Multiple Modified U-Nets and SegNets

Cited by 8 publications

References 25 publications

Hemorrhage semantic segmentation in fundus images for the diagnosis of diabetic retinopathy by using a convolutional neural network

Hemorrhage semantic segmentation in fundus images for the diagnosis of diabetic retinopathy by using a convolutional neural network

A deeply supervised maximum response texton basedSegNetfor simultaneous multi retinal lesion segmentation

Neighbored-attention U-net (NAU-net) for diabetic retinopathy image segmentation

Contact Info

Product

Resources

About