Clinical information prompt-driven retinal fundus image for brain health evaluation

The structure of the optic disc is impacted by the devastating eye disease glaucoma. Most doctors rely on viewing and screening the retinal fundus images to detect it. Identifying the optic disc is a crucial stage for automated diagnosis of glaucoma. This study presents a transfer learning-based approach to locate the optic disc in the retinal fundus image. The fundus images are categorized into four distinct classes according to the positions of optic disc to learn the in-depth specific features.Based on the YOLOv3 model, the optic disc detection method is then fine-tuned using the specialised image dataset. The model involved is trained and evaluated on 2178 retinal fundus images obtained from database at a nearby hospital and publically accessible datasets. In 2103 of the retinal fundus images, the suggested approach successfully detected the optical disc, yielding an average accuracy of 96.56%.

Heart disease represents a significant public health problem worldwide, and early discovery is required in order to improve clinical outcomes and decrease mortality. Retinal imaging can provide a noninvasive methodology for evaluating microvascular health, with strong associations to cardiovascular disease. This work introduces a simple deeplearning method to predict heart disease from retinal fundus images. The proposed system consists of preprocessing retinal fundus images, then extracting features with convolutional neural networks (CNN), and finally predicting heart disease using binary classification. The model displays strong predictive capabilities, demonstrating that retinal vascular patterns contain important cardiovascular risk biomarkers. The results suggest that retinal images could be an effective, scalable screening tool for early detection of heart disease.

The macula is the part of the eye responsible for central high acuity vision. Detection of the macula is an important task in retinal image processing as a landmark for subsequent disease assessment, such as for age-related macula degeneration. In this paper, we have presented an approach to automatically determine the macula centre in retinal fundus images. First contextual information on the image is combined with a statistical model to obtain an approximate macula region of interest localization. Subsequently, we propose the use of a seeded mode tracking technique to locate the macula centre. The proposed approach is tested on a large dataset composed of 482 normal images and 162 glaucoma images from the ORIGA database and an additional 96 AMD images. The results show a ROI detection of 97.5%, and 90.5% correct detection of the macula within 1/3DD from a manual reference, which outperforms other current methods. The results are promising for the use of the proposed approach to locate the macula for the detection of macula diseases from retinal images.

Using Deep Learning Models to Characterize Major Retinal Features on Color Fundus Photographs

This study focuses on detecting and classifying neovascularization caused by proliferative diabetic retinopathy (PDR) in retinal fundus images. Image processing methods were applied to detect retinal vessels. A fractal analysis approach based on the box-counting method was used to quantify vascular patterns in normal and abnormal cases showing neovascularization near the optic disk (NVD). Ten images including five normal cases and five neovascularization cases were analyzed. The mean fractal dimension obtained for the NVD cases was 1.66 compared to the mean value of 1.52 for the normal cases. The statistical significance of the difference was high, with a p-value of 0.0088. The results show promise for use in detecting neovascularization in retinal images caused by PDR.

The aim of this survey is to list the various disease predictions from retinal fundus images and various methods used to detect the disease. This paper gives a detailed description about the various diseases predicted in retina by comparing retinal fundus image structure. Till now, the prediction of various diseases such as diabetic retinopathy, cardiovascular disease and other eye problems had been predicted by using retinal fundus images. Next, a comparitive study of the various methods followed using image processing to find out the diseases from retinal fundus images, is provided. The basic matrices observed to predict the diseases are optic disc,nerve cup and rim. To find the differences in the basic matrices, image processing techniques such as mask generation, colour normalization, edge detection, contrast enhancement are used. The datasets that are used for retinal image inputs are STARE, DRIVE, ONHSD, ARIA, IMAGERET. The survey at the end, discusses the future work for the possibilities of predicting gastreointestinal problems via retinal fundus images.

In this study, we propose a CNN-GAN-based real-time processing technique for filtering images of underwater cables used in power systems. This addresses the excessive interference impurities that are frequently observed in images captured by remotely operated vehicles (ROVs). The process begins with the input of the original image into the convolutional neural network (CNN). Subsequently, the training outcomes, which serve as input parameters for the generative adversarial network (GAN), facilitate the filtering process. The system also calculates both the structural similarity index measure (SSIM) and peak signal-to-noise ratio (PSNR), performing model updates via backward propagation. This technique utilizes deep learning technologies to achieve rapid, real-time filtering of underwater cable images. The experimental results reveal that the loss function of the CNN reaches 0.16 with an accuracy of 97.5%, while the loss function of the adversarial GAN network approaches 0.05. Compared with traditional methods such as DDN, JORDOR, RESCAN, and PRENet, the proposed CNN-GAN algorithm exhibits superior performance, as evidenced by the higher PSNR and SSIM values. Specifically, for clear water images, the PSNR reaches 29.86 dB and the SSIM is 0.9045. For severely polluted images, the PSNR is 28.67 dB and the SSIM is 0.8965, while for unevenly illuminated images, the PSNR and SSIM values are 24.37 dB and 0.88, respectively. These enhancements significantly benefit the monitoring and maintenance of power systems.

PurposeTo assess the performance of deep learning algorithms for different tasks in retinal fundus images: (1) detection of retinal fundus images versus optical coherence tomography (OCT) or other images, (2) evaluation of good quality retinal fundus images, (3) distinction between right eye (OD) and left eye (OS) retinal fundus images,(4) detection of age-related macular degeneration (AMD) and (5) detection of referable glaucomatous optic neuropathy (GON).Patients and MethodsFive algorithms were designed. Retrospective study from a database of 306,302 images, Optretina’s tagged dataset. Three different ophthalmologists, all retinal specialists, classified all images. The dataset was split per patient in a training (80%) and testing (20%) splits. Three different CNN architectures were employed, two of which were custom designed to minimize the number of parameters with minimal impact on its accuracy. Main outcome measure was area under the curve (AUC) with accuracy, sensitivity and specificity.ResultsDetermination of retinal fundus image had AUC of 0.979 with an accuracy of 96% (sensitivity 97.7%, specificity 92.4%). Determination of good quality retinal fundus image had AUC of 0.947, accuracy 91.8% (sensitivity 96.9%, specificity 81.8%). Algorithm for OD/OS had AUC 0.989, accuracy 97.4%. AMD had AUC of 0.936, accuracy 86.3% (sensitivity 90.2% specificity 82.5%), GON had AUC of 0.863, accuracy 80.2% (sensitivity 76.8%, specificity 83.8%).ConclusionDeep learning algorithms can differentiate a retinal fundus image from other images. Algorithms can evaluate the quality of an image, discriminate between right or left eye and detect the presence of AMD and GON with a high level of accuracy, sensitivity and specificity.

Diabetic retinopathy (DR) is a complication of retinal related common micro vascular disease occurred in diabetic patients. It is the one of the major causes of adult blindness in the worldwide. DR is indicated by the presence of hard exudates, soft exudates, haemorrhages, the growth of new blood vessels and foveal avascular zone (FAZ) enlargement. The objective of this work is to detect FAZ area in the colour retinal fundus images. In retinal fundus images, FAZ is located in the centre of macula, which is usually capillary-free. The detection of FAZ is an important step in the assessment of retinal images for visual related pathology detection. Retinal images are taken from Messidor database consisting of 59 retinal images. The proposed method has four main steps, i.e. pre-processing, blood vessels extraction, FAZ detection, and validation test. In the pre-processing step, morphological operation followed by contrast stretching is used to enhance and improve image quality. In other hand, matched filter and local entropy thresholding are applied to extract retinal blood vessels. FAZ area is detected by connecting retinal vessel endpoints. Finally, validation test is conducted by calculating the correlation between polygon area as detected FAZ by proposed method and circular area as labelled FAZ by ophthalmologists. The result achieves the correlation coefficient of 0.912 indicating that the proposed method has significant correlation with the method of the ophthalmologists.

Non-invasive biopsy diagnosis of diabetic kidney disease via deep learning applied to retinal images: a population-based study.

In this research, retinal blood vessel fundus images will be segmented. Moreover, bifurcation point of the segmented blood vessel images will be determined. The data used for this research was obtained from fundus image database of DRIVE (Digital Retinal Images for Vessel Extraction) dataset. The preprocessing phases contain of several steps such as green channel extraction, histogram equalization and optic disc elimination. Meanwhile, the segmentation phase was performed by combining two filtering methods. These methods are median and derivative of Gaussian filter. Median filter was used to reduce the noise on retinal fundus image such as hard and soft exudates as well as dot and blot hemorrhages. Combination of these two filtering methods were also employed to determine the bifurcation point of segmented blood vessel image. The bifurcation points and the result from blood vessel segmentation were among two of the unique parameters of retinal fundus image. These parameters will be used eventually as unique features of the advanced research which is a biometric system to identify an individual based on his/her unique retinal pattern.

Artificial intelligence using deep learning to screen for referable and vision-threatening diabetic retinopathy in Africa: a clinical validation study

Retinal fundus images are commonly used by ophthalmologists to diagnose and monitor the development of eye diseases. One of the key information in the fundus images is to distinguish between the left and right eye. This work is motivated to provide an accurate and automated method to detect the left and right eye based on the retinal images.In current clinical practices, a common method to differentiate left and right eye in the retinal images is via the position of the optic nerve head with respect to the macula. This paper introduces a novel method to automatically detect the orientation of the retina in fundus images from the position of the central retinal artery and veins of the physiologic optic nerve head.Our method first locates the centre of the optic cup using pixel intensity. A 100 x 100 pixel ROI is then formed based on this centriod. Thresholding by pixel intensity and morphology is then used to segment the optic disc from this ROI to obtain an accurate optic disc and its blood vessels within. The pixel intensity in the green channel of this segmented disc along its horizontal axis is projected. The sum of the pixel intensity in the left half of the disc is then compared to the right half. The prevalence of higher green pixel intensity in either half of the optic disc would then indicate if this is a nasal or temporal sector, which would then indicate if this is a left or right eye.This method is tested on 194 images collected from patients at the Singapore Eye Research Institute and achieves an accuracy of 92.23%. This method is novel and its results are encouraging to provide an accurate and automatic system to make a distinction between left and right eye from retinal images.KeywordsAutomatedDetectionRetinalPixel Intensity

The research that used digital image processing to detect glaucoma has been known as one of the popular methods. The beginning step of the research is to choose the best channel among red, green, or blue (RGB) so it can ease the glaucoma segmentation. In choosing the channel, it is important to analyze deeply the used retinal images. Choosing of best component can affect the accuracy of glaucoma diagnosis results. In this research, the most suitable component will be analyzed to detect glaucoma based on the visual, MSE (Mean Square Error) value, and PSNR (Peak Signal to Noise Ratio). In this research, we used 85 images of glaucoma from the DRISTHI-GS database and 101 normal images from the RIM-ONE database. From the visualization, it showed that the red component had high brightness level so it can differ the optic disc and other parts of retinal eyes. The green component still has vessel blood so it will make it more difficult to segment the images. Blue component results in very dark of retinal image. From MSE and PNSR values, it showed that the green component had the smallest MSE value while the blue component has the biggest MSE value. PSNR value was obtained from the green component. Both red and blue components had PSNR value which had a small difference. From these results, it can be concluded that the MSE and PSNR values do not guarantee visual results. So that for further research, it is expected that the MSE and PSNR values will be obtained from the part that we want to observe

Purpose. To automate the retinal screening process and to improve chances of early detection/diagnosis of diabetic retinopathy which may reduce vision impairment cases due to the diseaseMethods. Experiments were conducted on EYEPACS dataset that had retinal fundus images of patients with diabetic retinopathy in all severity levels. Based on the learning done by the Inception networks during the training phase, the images were classified to 5 classes based on the severity as: Normal, Mild, Moderate, severe and proliferative diabetic retinopathy. The images were augmented to increase the number of images with the disease and were randomly chosen for the training, validation and test datasets.Results: The training was conducted using the Inception network for 100 epochs and validation accuracy was tested every 1000 iterations on a total of 109,540 images. The major misclassifications that occurred were between the adjacent classes of severity. It achieved a sensitivity of 94.6%, specificity of 81.4%, accuracy of 92% and precision of 93.2%. Conclusion: The retinal fundus images were examined and automated using Inception network by training the images from scratch. This model can be applied for detection of other retinal diseases, as the model weights are learnt to recognize subtle features present in the retinal images