Wyniki wyszukiwania - BazTech

1

Automated detection of crystalline retinopathy via fundus photography using multistage generative adversarial networks

Young Choi Eun, Han Seung Hoon, Hee Ryu Ik, Kuk Kim Jin, Sik Lee In, Han Eoksoo, Kim Hyungsu, Yul Choi Joon, Keun Yoo Tae

Biocybernetics and Biomedical Engineering

|

2023

|

Vol. 43, no. 4

725--735

EN

Purpose: Crystalline retinopathy is characterized by reflective crystal deposits in the macula and is caused by various systemic conditions including hereditary, toxic, and embolic etiologies. Herein, we introduce a novel application of deep learning with a multistage generative adversarial network (GAN) to detect crystalline retinopathy using fundus photography. Methods: The dataset comprised major classes (healthy retina, diabetic retinopathy, exudative age-related macular degeneration, and drusen) and a crystalline retinopathy class (minor set). To overcome the limited data on crystalline retinopathy, we proposed a novel multistage GAN framework. The GAN was retrained after CutMix combination by inputting the GAN-generated synthetic data as new inputs to the original training data. After the multistage CycleGAN augmented the data for crystalline retinopathy, we built a deep-learning classifier model for detection. Results: Using the multistage CycleGAN facilitated realistic fundus photography synthesis with the characteristic features of retinal crystalline deposits. The proposed method outperformed typical transfer learning, prototypical networks, and knowledge distillation for both multiclass and binary classifications. The final model achieved an area under the curve of the receiver operating characteristics of 0.962 for internal validation and 0.987 for external validation for the detection of crystalline retinopathy. Conclusion: We introduced a deep learning approach for detecting crystalline retinopathy, a potential biomarker of underlying systemic pathological conditions. Our approach enables realistic pathological image synthesis and more accurate prediction of crystalline retinopathy, an essential but minor retinal condition.

2

Devignetting fundus images via Bayesian estimation of illumination component and gamma correction

James Shine P., Abraham Chandy D.

Biocybernetics and Biomedical Engineering

|

2021

|

Vol. 41, no. 3

1071--1092

EN

Background: Fundus photography is an imaging modality exclusively used in ophthalmology for visualizing structures like macula, retina, and optic disc. The fundus camera has only one illumination source, which is situated at its center. Hence, structures away from the center will appear darker than naturally they are. This adverse effect caused by the uneven illumination is called as ‘vignetting’. Objectives: An algorithm termed as Gamma Correction of Illumination Component (GCIC) for devignetting fundus images is proposed in this paper. Methods: Inspired by the Retinex theory, the illumination component is computed with the help of a Maximum a Posteriori (MAP) estimator. The estimated illumination component after normalization is subjected to the Gamma correction to suppress its unevenness. Results: GCIC exhibited comparatively low values of Average Gradient of the Illumination Component (AGIC), Lightness Order Error (LOE), and computational time. The proposed method gave a comparatively better performance in terms of the performance metrics, namely contrast-to-noise-ratio (CNR), peak-signal-to-noise-ratio (PSNR), structure similarity index (SSIM), and entropy. With respect to the cumulative performance, GCIC has been observed to be better than other devignetting algorithms in the literature, like Illumination Equalization model, Homomorphic Filtering, Adaptive Gamma Correction (AGC), Modified Sigmoid Transform (MST), Imran Qureshi et al. (2019), Zheng et al., Variation-based Fusion (VF) and Zhou’s et al. Conclusion: GCIC corrects the uneven background illumination without scaling or boosting it intolerably. It produces output images, which are natural in appearance, free from color artefacts, and maintaining the sharpness of the fundus features. It is computationally fast as well.

3

Automated segmentation of optic disc and optic cup for glaucoma assessment using improved UNET++ architecture

Tulsani Akshat, Kumar Preetham, Pathan Sumaiya

Biocybernetics and Biomedical Engineering

|

2021

|

Vol. 41, no. 2

819--832

EN

Glaucoma is one of the leading cause of blindness for over 60 million people around the world. Since a cure for glaucoma doesn’t yet exist, early screening and diagnosis become critical for the prevention of the disease. Optic disc and optic cup evaluation are one of the preeminent steps for glaucoma diagnosis. A novel approach is developed in this paper for the identification of glaucoma using a segmentation based approach on the optic disc and optic cup. The Dhristi dataset was used to help improve performance on a small dataset. A custom UNET++ model is built for the segmentation task by tuning the hyperparameters in addition to a custom loss function. The developed loss function helps tackle the class imbalance occurring due to small size of the optic nerve head. The proposed approach achieves 96% accuracy in classifying glaucomatous and non-glaucomatous images based on clinical feature identification. The improvised model is able to achieve state-of-art results for Intersection over Union (IOU) scores, 0.9477 for optic disc and 0.9321 for optic cup, along with providing an enhancement in reducing the training time. The model was tested on publicly available datasets RIM-ONE, DRIONS-DB and ORIGA and is able to achieve an accuracy of 91%, 92% and 90% respectively. The developed approach is validated by training it over RIM-ONE dataset independently, without changing any model parameters. The model provides significant improvement in segmentation of the optic disc and optic cup along with improvement in training time.

4

Modified U-Net architecture for semantic segmentation of diabetic retinopathy images

Sambyal Nitigya, Saini Poonam, Syal Rupali, Gupta Varun

Biocybernetics and Biomedical Engineering

|

2020

|

Vol. 40, no. 3

1094--1109

EN

Segmentation of lesions from fundus images is an essential prerequisite for accurate severity assessment of diabetic retinopathy. Due to variation in morphologies, number and size of lesions, the manual grading process becomes extremely challenging and time-consuming. This necessitates the need of an automatic segmentation system that can precisely define the region of interest boundaries and assist ophthalmologists in speedy diagnosis along with diabetic retinopathy severity grading. The paper presents a modified U-Net architecture based on residual network and employs periodic shuffling with sub-pixel convolution initialized to convolution nearest neighbour resize. The proposed architecture has been trained and validated for microaneurysm and hard exudate segmentation on two publicly available datasets namely IDRiD and e-ophtha. For IDRiD dataset, the network obtains 99.88% accuracy, 99.85% sensitivity, 99.95% specificity and dice score of 0.9998 for both microaneurysm and exudate segmentation. Further, when trained on e-ophtha and validated on IDRiD dataset, the network shows 99.98% accuracy, 99.88% sensitivity, 99.89% specificity and dice score of 0.9998 for microaneurysm segmentation. For exudates segmen-tation, the model obtained 99.98% accuracy, 99.88% sensitivity, 99.89% specificity and dice score of 0.9999, when trained on e-ophtha and validated on IDRiD dataset. In comparison to existing literature, the proposed model provides state-of-the-art results for retinal lesion segmentation.

5

Automated detection of optic disc contours in fundus images using decision tree classifier

Pathan Sumaiya, Kumar Preetham, Pai Radhika, Bhandary Sulatha V.

Biocybernetics and Biomedical Engineering

|

2020

|

Vol. 40, no. 1

52--64

EN

Automated segmentation of optic disc in fundus images plays a vital role in computer aided diagnosis (CAD) of eye pathologies. In this paper, a novel method is proposed which detects and excludes the blood vessel for accurate optic disc segmentation. This is achieved in two steps. First, an effective blood vessel detection and exclusion algorithm is developed using directional filter. In the second step, a decision tree classifier is used to obtain an adaptive threshold in order to detect the contour of optic disc. The proposed method aids in computationally robust segmentation of optic disc even in fundus images having illuminations, reflections and exudates. The proposed method is tested on two different datasets which includes 300 fundus images collected from Kasturba Medical College (KMC) Manipal and also the publically available RIM-ONE database. The average values of Jaccard index, dice coefficient, sensitivity, specificity and accuracy obtained for KMC images is 91.28 %, 94.17 %, 92.71 %, 99.89 % and 99.61 % respectively. For RIM-ONE database the obtained average values of Jaccard index, dice coefficient, sensitivity, specificity and accuracy are 85.30 %, 90.69 %, 93.90 %, 99.39 % and 99.15 % respectively. The obtained segmentation results proves the efficiency of the algorithm to be incorporated in CAD of eye diseases.

6

A hybrid feature preservation technique based on luminosity and edge based contrast enhancement in color fundus images

Palanisamy Gopinath, Shankar Natarajan B., Ponnusamy Palanisamy, Gopi Varun P.

Biocybernetics and Biomedical Engineering

|

2020

|

Vol. 40, no. 2

752--763

EN

Color fundus image analysis for detecting the retinal abnormalities requires an improved visualization of image attributes with sufficient luminosity, contrast and accurate edge details. A hybrid technique based on singular value equalization using shearlet transform and adaptive gamma correction, followed by contrast limited adaptive histogram equalization (CLAHE) is proposed for the enhancement of luminosity and contrast in color fundus images. The low frequency components of the original and adaptive gamma transformed value channel in HSV color space obtained by applying shearlet transform are considered for singular value equalization. The high frequency components of the unchanged value channel, denoised using soft thresholding are applied while performing inverse shearlet transform. Luminosity component in Lab colorspace is considered for performing CLAHE on the singular value equalized image. Subjective analysis is done based on visualization of the image attributes and the objective analysis is carried out based on the parameters such as Peak signal to noise ratio, entropy, feature similarity index, edge-based contrast measure, quality index and noise suppression measure. The simulation results evince superior noise performance, sufficient luminosity adjustment and improved contrast along with excellent edge detail preservation when compared with the existing state-of-the-art techniques.

7

Estimation of severity level of non-proliferative diabetic retinopathy for clinical aid

Kaur J., Mittal D.

Biocybernetics and Biomedical Engineering

|

2018

|

Vol. 38, no. 3

708--732

EN

Diabetic retinopathy, a symptomless complication of diabetes, is one of the significant causes of vision impairment in the world. The early detection and diagnosis can reduce the occurrence of severe vision loss due to diabetic retinopathy. The diagnosis of diabetic retinopathy depends on the reliable detection and classification of bright and dark lesions present in retinal fundus images. Therefore, in this work, reliable segmentation of lesions has been performed using iterative clustering irrespective of associated heterogeneity, bright and faint edges. Afterwards, a computer-aided severity level detection method is proposed to aid ophthalmologists for appropriate treatment and effective planning in the diagnosis of non-proliferative diabetic retinopathy. This work has been performed on a composite database of 5048 retinal fundus images having varying attributes such as position, dimensions, shapes and color to make a reasonable comparison with state-of-the-art methods and to establish generalization capability of the proposed method. Experimental results on per-lesion basis show that the proposed method outperforms state-of-the methods with an average sensitivity/specificity/accuracy of 96.41/96.57/94.96 and 95.19/96.24/96.50 for bright and dark lesions respectively on composite database. Individual per-image based class accuracies delivered by the proposed method: No DR-95.9%, MA-98.3%, HEM-98.4%, EXU-97.4% and CWS-97.9% demonstrate the clinical competence of the method. Major contribution of the proposed method is that it efficiently grades the severity level of diabetic retinopathy in spite of huge variations in retinal images of different databases. Additionally, the substantial combined performance of these experiments on clinical and open source benchmark databases support a strong candidature of the proposed method in the diagnosis of non-proliferative diabetic retinopathy.

8

Combination of clinical and multiresolution features for glaucoma detection and its classification using fundus images

Kausu T. R., Gopi V. P., Wahid K. A., Doma W., Niwas S. I.

Biocybernetics and Biomedical Engineering

|

2018

|

Vol. 38, no. 2

329--341

EN

Glaucoma is a neuro-degenerative disorder of the eye and it leads to permanent blindness when untreated or detected in the later stage. The main cause of glaucoma is the damage of the optic nerve, which occurs due to the increase of eye pressure. Hence the early detection of this disease is critical in time and which can help to prevent further vision loss. The assessment of optic nerve head using fundus images is more beneficial than the raised intra ocular pressure assessment in population-based glaucoma screening. This work proposed a novel method for glaucoma identification based on time-invariant feature cup to disk ratio and anisotropic dual-tree complex wavelet transform features. Optic disk segmentation is done by using Fuzzy C-Means clustering method and Otsu's thresholding is used for optic cup segmentation. The results show the proposed method achieved an accuracy rate of 97.67% with 98% sensitivity using a multilayer perceptron model that is considered as clinically significant when compared to the existing works.

9

Automated fuzzy optic disc detection algorithm using branching of vessels and color properties in fundus images

Nergiz M., Akın M., Yıldız A., Takeş Ö.

Biocybernetics and Biomedical Engineering

|

2018

|

Vol. 38, no. 4

850--867

EN

Optic disc (OD) detection is a basic procedure for the image processing algorithms which intend to diagnose and track retinal disorders. In this study, a new OD localization approach is proposed, based on color and shape properties of OD as well as the convergence point of the main vessels. This study is comprised of two successive fundamental steps. At the first step, an algorithm finding the approximate convergent point of the vessels is used in order to roughly localize OD. At the second step, three new features are suggested and a fuzzy logic controller (FLC) whose input membership functions are designed based on these features is proposed. The proposed method is applied to the DRIVE, STARE, DIARETDB0 and DIRETDB1 datasets and the obtained results validate the improvement in the performance by attaining success rate of 100%, 91,35%, 90% and 100% respectively and detecting OD centers and contours precisely in a reasonable execution time.

10

A generalized method for the segmentation of exudates from pathological retinal fundus images

Kaur J., Mittal D.

Biocybernetics and Biomedical Engineering

|

2018

|

Vol. 38, no. 1

27--53

EN

Diabetic retinopathy, an asymptomatic complication of diabetes, is one of the leading causes of blindness in the world. The exudates, abnormal leaked fatty deposits on retina, are one of the most prevalent and earliest clinical signs of diabetic retinopathy. In this paper, a generalized exudates segmentation method to assist ophthalmologists for timely treatment and effective planning in the diagnosis of diabetic retinopathy is developed. The main contribution of the proposed method is the reliable segmentation of exudates using dynamic decision thresholding irrespective of associated heterogeneity, bright and faint edges. The method is robust in the sense that it selects the threshold value dynamically irrespective of the large variations in retinal fundus images from varying databases. Since no performance comparison of state of the art methods is available on common database, therefore, to make a fair comparison of the proposed method, this work has been performed on a diversified database having 1307 retinal fundus images of varying characteristics namely: location, shapes, color and sizes. The database comprises of 649 clinically acquired retinal fundus images from eye hospital and 658 retinal images from publicly available databases such as STARE, MESSIDOR, DIARETDB1 and e-Optha EX. The segmentation results are validated by performing two sets of experiments namely: lesion based evaluation criteria and image based evaluation criteria. Experimental results at lesion level show that the proposed method outperforms other existing methods with a mean sensitivity/specificity/accuracy of 88.85/96.15/93.46 on a composite database of retinal fundus images. The segmentation results for image-based evaluation with a mean sensitivity/specificity/accuracy of 94.62/ 98.64/96.74 respectively prove the clinical effectiveness of the method. Furthermore, the significant collective performance of these experiments on clinically as well as publicly available standard databases proves the generalization ability and the strong candidature of the proposed method in the real-time diagnosis of diabetic retinopathy.

11

Retinal blood vessel segmentation employing image processing and data mining techniques for computerized retinal image analysis

GeethaRamani R., Balasubramanian L.

Biocybernetics and Biomedical Engineering

|

2016

|

Vol. 36, no. 1

102--118

EN

Most of the retinal diseases namely retinopathy, occlusion etc., can be identified through changes exhibited in retinal vasculature of fundus images. Thus, segmentation of retinal blood vessels aids in detecting the alterations and hence the disease. Manual segmentation of vessels requires expertise. It is a very tedious and time consuming task as vessels are only a few pixels wide and extend almost throughout entire span of the fundus image. Employing computational approaches for this purpose would help in efficient retinal analysis. The methodology proposed in this work involves sequential application of image pre-processing, supervised and unsupervised learning and image post-processing techniques. Image cropping, color transformation and color channel extraction, contrast enhancement, Gabor filtering and halfwave rectification are sequentially applied during pre-processing stage. A feature vector is formed from the pre-processed images. Principal component analysis is performed on the feature vector. K-means clustering is executed on this outcome to group pixels as either vessel or non-vessel cluster. Out of the two groups, the identified non-vessel group undergoes an ensemble classification process employing root guided decision tree with bagging, while vessel group is left unprocessed as further processing might increase misclassifications of vessels as non-vessels. The resultant segmented image is formed through combining the results of clustering and ensemble classification process. The vessel segmented output from previous phase is post-processed through morphological techniques. The proposed technique is validated on images from publicly available DRIVE database. The proposed methodology achieves an accuracy of 95.36%, which is comparable with the existing blood vessel segmentation techniques.