Abnormality detection in retinal image by individualized background learning

Chen, Benzhi; Wang, Lisheng; Wang, Xiuying; Sun, Jian; Huang, Yalin; Feng, David Dagan; Zhao, Xu

doi:10.1016/j.patcog.2020.107209

Cited by 10 publications

(4 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then a morphology method is applied to identify H, MA, and EX [5]. Similarly, Chen et al, preprocess the images by suppressing the retinal blood vessels and use a multi-scale sparse coding-based learning algorithm to learn the individualized retinal background and identify salient lesions [13]. Alternatively, Amin et al, present the combination of Gabor filter mathematical morphology, statistical and geometric features to detect EX and to grade DR using different ensembles of classifiers [6].…”

Section: Diabetic Retinopathy-related Lesions Detectionmentioning

confidence: 99%

A Benchmark for Studying Diabetic Retinopathy: Segmentation, Grading, and Transferability

Zhou

Wang

Huang

et al. 2021

IEEE Trans. Med. Imaging

139

View full text Add to dashboard Cite

Diabetic retinopathy (DR) is the result of a complication of diabetes affecting the retina. It can cause blindness if left undiagnosed and untreated. The ophthalmologist performs the diagnosis by screening each patient and detecting in ocular imaging the lesions caused by DR, namely, microaneurisms, hemorrhages, cotton wool spots, venous beading and neovascularization. However, the analysis of ocular findings is cumbersome, time-consuming, and demanding. Due to the insufficient amount of trained specialists to diagnose the illness, and the actual growing population with DR, it is important to develop a method to assist the DR diagnosis. This thesis presents two approaches for the automatic classification of DR using eye fundus images. The first one utilizes convolutional neural networks, transfer learning and shallow machine learning classifiers to identify the main ocular lesions related to DR and then use them to diagnose the illness. The second one is a multitask model which predicts simultaneously ocular lesions and DR. These approaches follow a similar workflow to that of clinicians, providing information that can be interpreted clinically to support the prediction. To achieve this goal a subset of the kaggle EyePACS and the Messidor-2 datasets, are labeled with ocular lesions by a certified opthalmologist. The kaggle EyePACS subset is used as training set and the Messidor-2 dataset is used as test set for both, the lesions and DR classification models. The results indicate that both methods achieve results comparable with state-of-the-art performances. The best results are obtained using the first approach with a multi layer perceptron as classifier for the automatic detection of DR, however, the multitask approach lead to similar results and has a simpler architecture.

show abstract

Section: Diabetic Retinopathy-related Lesions Detectionmentioning

confidence: 99%

A Benchmark for Studying Diabetic Retinopathy: Segmentation, Grading, and Transferability

Zhou

Wang

Huang

et al. 2021

IEEE Trans. Med. Imaging

139

View full text Add to dashboard Cite

show abstract

“…In the diagnosis and evaluation of diseases, the accurate segmentation of the MA regions requires high standards. If the MA is segmented manually, the segmentation results will be subjectively affected and manual segmentation is also time‐consuming 9 . Therefore, it is desirable to develop AI algorithms to improve the accuracy and efficiency of MA region segmentation.…”

Section: Introductionmentioning

confidence: 99%

“…If the MA is segmented manually, the segmentation results will be subjectively affected and manual segmentation is also time-consuming. 9 Therefore, it is desirable to develop AI algorithms to improve the accuracy and efficiency of MA region segmentation. In the past, researchers have developed many algorithms for automatic MA segmentation.…”

mentioning

confidence: 99%

Microaneurysm classification system in color fundus images using auto‐weight dilated convolutional network

Jayachandran,

Ratheesh Kumar

2023

Int J Imaging Syst Tech

View full text Add to dashboard Cite

Microaneurysms (MAs) are the early indications of diabetic retinopathy (DR), which may result in total visual loss. MAs detection is an exciting work due to its small, darkish color and subtle nature. The automatic detection and categorization of MAs in retinal fundus images using a multi‐scale approach based on Deep Neural Networks and Neighborhood Analysis is proposed in this research article. MAs segmentation, classification, and preprocessing comprise the three steps of the proposed technique. To extract the multi‐scale MA features, the auto‐weight dilated convolutional unit (AD) is specifically used for convolutional feature maps. To fuse convolutional feature maps in encoding layers, the AD unit used a learnable set of parameters. An efficient architecture for feature extraction during the encoding step is incorporated into the AD unit. We integrated the AD unit with the benefits of the U‐Net network for deep and shallow features. Additionally, in order to improve the suggested model and produce the final derivation, we developed a novel optimization approach. After that, neighborhood analysis is performed to name the Micro‐aneurysm because the lesion is actually a collection of independent little images rather than the entire image. The classification accuracy of the proposed method for the three different data sets such as MESSIDOR, online retinal fundus image database for glaucoma analysis (ORIGA), and RIM‐ONE‐R1, is 99.28%, 98.95%, and 98.76% respectively. The results show a good performance of the proposed model against the other analyzed procedure.

show abstract

“…The change regions are modelled as the foreground objects and obtained through subtracting the learned background from the current frames 14 , 15 . Principal component analysis (PCA) extracts the most of similar content linearly from a set of images 12 , 16 or a serial 7 , 17 as the background model and filters the noise or disturbance.…”

Section: Introductionmentioning

confidence: 99%

Change detection based on unsupervised sparse representation for fundus image pair

Zhao

Liang

et al. 2022

Sci Rep

View full text Add to dashboard Cite

Detecting changes is an important issue for ophthalmology to compare longitudinal fundus images at different stages and obtain change regions. Illumination variations bring distractions on the change regions by the pixel-by-pixel comparison. In this paper, a new unsupervised change detection method based on sparse representation classification (SRC) is proposed for the fundus image pair. First, the local neighborhood patches are extracted from the reference image to build a dictionary of the local background. Then the current image patch is represented sparsely and its background is reconstructed by the obtained dictionary. Finally, change regions are given through background subtracting. The SRC method can correct automatically illumination variations through the representation coefficients and filter local contrast and global intensity effectively. In experiments of this paper, the AUC and mAP values of SRC method are 0.9858 and 0.8647 respectively for the image pair with small lesions; the AUC and mAP values of the fusion method of IRHSF and SRC are 0.9892 and 0.9692 separately for the image pair with the big change region. Experiments show that the proposed method in this paper is more robust than RPCA for the illumination variations and can detect change regions more effectively than pixel-wised image differencing.

show abstract

Abnormality detection in retinal image by individualized background learning

Cited by 10 publications

References 29 publications

A Benchmark for Studying Diabetic Retinopathy: Segmentation, Grading, and Transferability

A Benchmark for Studying Diabetic Retinopathy: Segmentation, Grading, and Transferability

Microaneurysm classification system in color fundus images using auto‐weight dilated convolutional network

Change detection based on unsupervised sparse representation for fundus image pair

Contact Info

Product

Resources

About