Restoration of retinal images obtained through cataracts

Peli, Eli; Peli, Tamar

doi:10.1109/42.41493

Cited by 47 publications

(19 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…From a macro point of view, the composition of the CAD system is very similar to the process of early clinical diagnosis and screening of fundus diseases by ophthalmologists by observing fundus images. If the whole process of early screening methods is compared to an assembly line, then at the front end of the pipeline, CAD can automatically perceive the distortion of the fundus image [13] and perform corresponding image enhancement [14], [36] and restoration [15], [35]; In the intermediate stage, CAD can segment the lesion and extract many features; At the back end , CAD can transform complex diagnostic logic into classification or clustering problems in machine logic, and classification or clustering problems can be solved by machine learning. Like CAD, image preprocessing and then image segmentation are performed [16], [17], then feature extraction, and finally the process of intelligent diagnosis through machine learning is actually imitating the doctor's diagnostic thinking.…”

Section: Related Work a Traditional Methodsmentioning

confidence: 99%

Multi-Label Classification of Fundus Images With EfficientNet

et al. 2020

View full text Add to dashboard Cite

Convolutional neural network (CNN) has achieved remarkable success in the field of fundus images due to its powerful feature learning ability. Computer-aided diagnosis can obtain information with reference value for doctors in clinical diagnosis or screening through proper processing and analysis of fundus images. However, most of the previous studies have focused on the detection of a certain fundus disease, and the simultaneous diagnosis of multiple fundus diseases still faces great challenges. We propose a multi-label classification ensemble model of fundus images based on CNN to directly detect one or more fundus diseases in the retinal fundus images. Every single model consists of two parts. The first part is a feature extraction network based on EfficientNet, and the second part is a custom classification neural network for multi-label classification problems. Finally, the output probabilities of different models are fused as the final recognition result. And it was trained and tested on the data set provided by ODIR 2019 (Peking University International Competition on Ocular Disease Intelligent Recognition). The experimental results show that our model can be trained on fewer data sets and get good results.

show abstract

Section: Related Work a Traditional Methodsmentioning

confidence: 99%

Multi-Label Classification of Fundus Images With EfficientNet

et al. 2020

View full text Add to dashboard Cite

show abstract

“…During the final stage, the weights and the bias are updated by using the δ factor and the activation, by using α as the learning rate. Each output unit updates its weights and the bias by using the weight correction term ΔW jk and the bias correction term ΔW ok , where ΔW jk = αδ k z j and (18) Also, the weights and the bias are updated for each hidden unit. Finally, the stopping condition was tested, where the stopping condition may be a minimisation of the errors, the number of epochs and so on.…”

Section: Classificationmentioning

confidence: 99%

“…Cree et al [17] have applied an image restoration technique for images of very poor quality. Peli and Peli [18] have also applied the image restoration technique and in this approach, the images were considered as sample functions of stochastic process. Hsu et al [19] have proposed a technique to find the normal and the abnormal areas of the retinal images by using intensity properties for dynamic clustering.…”

Section: Introductionmentioning

confidence: 99%

Diagnosis of diabetic retinopathy by employing image processing technique to detect exudates in retinal images

Franklin

Rajan²

2014

IET Image Processing

View full text Add to dashboard Cite

Diabetic retinopathy (DR) is a microvascular complication of long-term diabetes and it is the major cause of visual impairment because of changes in blood vessels of the retina. Major vision loss because of DR is highly preventable with regular screening and timely intervention at the earlier stages. The presence of exudates is one of the primitive signs of DR and the detection of these exudates is the first step in automated screening for DR. Hence, exudates detection becomes a significant diagnostic task, in which digital retinal imaging plays a vital role. In this study, the authors propose an algorithm to detect the presence of exudates automatically and this helps the ophthalmologists in the diagnosis and follow-up of DR. Exudates are normally detected by their high grey-level variations and they have used an artificial neural network to perform this task by applying colour, size, shape and texture as the features. The performance of the authors algorithm has been prospectively tested by using DIARETDB1 database and evaluated by comparing the results with the ground-truth images annotated by expert ophthalmologists. They have obtained illustrative results of mean sensitivity 96.3%, mean specificity 99.8%, using lesion-based evaluation criterion and achieved a classification accuracy of 99.7%.

show abstract

“…coeff. = (2) LLf/(x,y) JC y where f 1 (x,y) and f 2 (x,y) are the two images, and the double summations are performed over all pixels that have valid data in both images after the transformation.…”

Section: L T(xy) H(xy)mentioning

confidence: 99%

<title>Registration of high-resolution images of the retina</title>

Cideciyan

Jacobson

Kemp

et al. 1992

Medical Imaging VI: Image Processing

View full text Add to dashboard Cite

A method of image registration is presented for the case when the deformation between two images can be well approximated with a combination of translation, rotation and global scaling. The method achieves very high accuracy by combining a global optimization in the 4-dimensional discrete parameter space with a local optimization in the 4-dimensional continuous parameter space. The 4-dimensional global optimization is accomplished with two 2-dimensional optimizations. The Fourier magnitude is used to decouple translation from rotation and scaling, and a log-polar mapping of the Fourier magnitude is used to convert rotation and scaling into shifts. Optimal rotation and scaling parameters are determined with a cross-correlation in the log-polar domain. After compensation for rotation and scaling differences, cross-correlation in the spatial domain yields the translation parameters. The four registration parameters are further refined with a local optimization using the correlation coefficient as a similarity measure in the 4-dimensional continuous parameter space. Results are shown from simulations and from registration of retinal images. For simulated images with a signal-to-noise ratio of -5 dB, the accuracy of the registration method is estimated 10 be better than 0.07 degrees, 0.1 %, and 0.3 pixels for rotation, scaling, and translation, respectively. In the case of 512x512 pixel images the computation resource requirements are compatible with high end PCs, i.e., approximately 25 minutes on an Intel 80486(33MHz based IBM/PC compatible.

show abstract

Restoration of retinal images obtained through cataracts

Cited by 47 publications

References 12 publications

Multi-Label Classification of Fundus Images With EfficientNet

Multi-Label Classification of Fundus Images With EfficientNet

Diagnosis of diabetic retinopathy by employing image processing technique to detect exudates in retinal images

<title>Registration of high-resolution images of the retina</title>

Contact Info

Product

Resources

About