Imaging Of Human Retina Research Articles

Computational Approach for Detection of Diabetes from Ocular Scans.

The estimated 30 million children and adults are suffering with diabetes across the world. A person with diabetes can recognize several symptoms, and it can also be tested using retina image as diabetes also affects the human eye. The doctor is usually able to detect retinal changes quickly and can help prevent vision loss. Therefore, regular eye examinations are very important. Diabetes is a chronic disease that affects various parts of the human body including the retina. It can also be considered as major cause for blindness in developed countries. This paper deals with classification of retinal image into diabetes or not with the help of deep learning algorithms and architecture. Hence, deep learning is beneficial for classification of medical images specifically such a complex image of human retina. A large number of image data are considered throughout the project on which classification is performed by using binary classifier. On applying certain deep learning algorithms, model results into the training accuracy of 96.68% and validation accuracy of 66.82%. Diabetic retinopathy can be considered as an effective and efficient method for diabetes detection.

OCT and OCT Angiography Offer New Insights and Opportunities in Schizophrenia Research and Treatment.

The human retina and retinal imaging technologies continue to increasingly gain the attention of schizophrenia researchers. With the same embryologic origin as the brain, the retina offers a window into neurovascular changes that may underlie disease. Recently, two technologies that have already revolutionized the field of ophthalmology, optical coherence tomography (OCT), and a functional extension of this, optical coherence tomography angiography (OCTA), have gained traction. Together, these non-invasive technologies allow for microscopic imaging of both structural and vascular features of the retina. With ease of use and no side effects, these devices are likely to prove powerful digital health tools in the study and treatment of schizophrenia. They may also prove key to discovering disease relevant biomarkers that underly neurodevelopmental and neurodegenerative aspects of conditions such as schizophrenia.

An Effective Approach Based on Improved Convolutionary Neural Network Classifier for the Detection of Diabetic Retinopathy

In diabetic individuals, diabetic retinopathy (DR) causes blindness. Therefore, detecting diabetic retinopathy at an early stage decreases vision loss. An successful approach for diabetic retinopathy prediction is discussed in this article. In the beginning, the input pictures of human retinal fundus images are preprocessed using histogram equalisation followed by Gabor filtering to reduce noise for enhancement. Then, using the Watershed method, segmentation is performed, and the features are retrieved through feature extraction. The best optimum features are selected using PCA (principal component analysis) approach. The morphological based post processing scheme was employed to further enhance the quality of selected features. At last, the classification approach is carried with the utilization of Google NET CNN classifier to classify/predict the retinal image as normal, abnormal, and severe. Google NET CNN has been developed with limited preprocessing step to distinguish visual features directly from image pixels. The findings are then evaluated and the efficacy of the new method is contrasted with other current methods. The quantitative findings were evaluated for Accuracy, precision, reliability, positive predictive levels and false predictive levels in parameters and were seen to deliver better results than current techniques.

Functional imaging of human retina using integrated multispectral and laser speckle contrast imaging.

A novel integration of retinal multispectral imaging (MSI), retinal oximetry and laser speckle contrast imaging (LSCI) is presented for functional imaging of retinal blood vessels that could potentially allow early detection or monitoring of functional changes. We designed and built a cost-effective, scalable, retinal imaging instrument that integrates structural and functional retinal imaging techniques, including MSI, retinal oximetry and LSCI. Color fundus imaging was performed with 470 nm, 550 nm and 600 nm wavelength light emitting diode (LED) illumination. Retinal oximetry was performed using 550 nm and 600 nm LED illumination. LSCI of blood flow was performed using 850 nm laser diode illumination at 82 frames per second. LSCI can visualize retinal and choroidal vasculature without requiring exogenous contrast agents and can provide time-resolved information on blood flow, generating a cardiac pulse waveform from retinal vasculature. The technology can rapidly acquire structural MSI images, retinal oximetry and LSCI blood flow information in a simplified clinical workflow without requiring patients to move between instruments. Results from multiple modalities can be combined and registered to provide structural as well as functional information on the retina. These advances can reduce barriers for clinical adoption, accelerating research using MSI, retinal oximetry and LSCI of blood flow for diagnosis, monitoring and elucidating disease pathogenesis.

In vivo human retinal swept source optical coherence tomography and angiography at 830\xa0nm with a CMOS compatible photonic integrated circuit

Photonic integrated circuits (PIC) provide promising functionalities to significantly reduce the size and costs of optical coherence tomography (OCT) systems. This paper presents an imaging platform operating at a center wavelength of 830 nm for ophthalmic application using PIC-based swept source OCT. An on-chip Mach–Zehnder interferometer (MZI) configuration, which comprises an input power splitter, polarization beam splitters in the sample and the reference arm, and a 50/50 coupler for signal interference represents the core element of the system with a footprint of only (12 times 5);{text {mm}}^2. The system achieves 94 dB imaging sensitivity with 750 upmu W on the sample, 50 kHz imaging speed and 5.5 upmu m axial resolution (in soft tissue). With this setup, in vivo human retinal imaging of healthy subjects was performed producing B-scans, three-dimensional renderings as well as OCT angiography. These promising results are significant prerequisites for further integration of optical and electronic building blocks on a single swept source-OCT PIC.

An Effective Approach Based on Deep Residual Google Net Convolutional Neural Network Classifier for the Detection of Glaucoma

Glaucoma is a disease caused by fluid pressure build-up in the inner eye. Early detection of glaucoma is critical as it is expected that 111.8 million people worldwide shall suffer from glaucoma in 2040. In the diagnosis of glaucoma, the use of machine learning method is hoped to be highly promising. This paper provides an important method to master learning to diagnose glaucoma. Initially, human retinal fundus images are preprocessed by means of histogram equalization in order to enhance them. The segmentation is performed by semantic segmentation method, mainly the features are extracted using density with correlation based feature extraction approach. PCA (principal component analysis) methodology is used to choose the most optimal features. Ultimately, through the usage of the Deep residual Google Net CNN Classification method, the retinal image is classified/predicted as regular and abnormal. The Deep residual Google Net CNN classifier is designed to distinguish view patterns with minimal pre-processing from pixel pictures. ORIGA and STARE datasets are used in this work. The findings are then analyzed and contrasted to illustrate the efficacy of the new technique with alternate current techniques. Test accuracy of 99%, Specificity of 98.9% and 100% Sensitivity were achieved. The quantitative results are analyzed for specifications like sensitivity, specificity, accuracy, positive predictive rate, false predictive rate and assured to provide excellent outcomes when compared with traditional methods.

Double-path parallel convolutional neural network for removing speckle noise in different types of OCT images

Speckle noises widely exist in optical coherence tomography (OCT) images. We propose an improved double-path parallel convolutional neural network (called DPNet) to reduce speckles. We increase the network width to replace the network depth to extract deeper information from the original OCT images. In addition, we use dilated convolution and residual learning to increase the learning ability of our DPNet. We use 100 pairs of human retinal OCT images as the training dataset. Then we test the DPNet model for denoising speckles on four different types of OCT images, mainly including human retinal OCT images, skin OCT images, colon crypt OCT images, and quail embryo OCT images. We compare the DPNet model with the adaptive complex diffusion method, the curvelet shrinkage method, the shearlet-based total variation method, and the OCTNet method. We qualitatively and quantitatively evaluate these methods in terms of image smoothness, structural information protection, and edge clarity. Our experimental results prove the performance of the DPNet model, and it allows us to batch and quickly process different types of poor-quality OCT images without any parameter fine-tuning under a time-constrained situation.

Multi-scale GCN-assisted two-stage network for joint segmentation of retinal layers and discs in peripapillary OCT images.

An accurate and automated tissue segmentation algorithm for retinal optical coherence tomography (OCT) images is crucial for the diagnosis of glaucoma. However, due to the presence of the optic disc, the anatomical structure of the peripapillary region of the retina is complicated and is challenging for segmentation. To address this issue, we develop a novel graph convolutional network (GCN)-assisted two-stage framework to simultaneously label the nine retinal layers and the optic disc. Specifically, a multi-scale global reasoning module is inserted between the encoder and decoder of a U-shape neural network to exploit anatomical prior knowledge and perform spatial reasoning. We conduct experiments on human peripapillary retinal OCT images. We also provide public access to the collected dataset, which might contribute to the research in the field of biomedical image processing. The Dice score of the proposed segmentation network is 0.820 ± 0.001 and the pixel accuracy is 0.830 ± 0.002, both of which outperform those from other state-of-the-art techniques.

A Comparison of Artificial Intelligence and Human Diabetic Retinal Image Interpretation in an Urban Health System.

Artificial intelligence (AI) diabetic retinopathy (DR) software has the potential to decrease time spent by clinicians on image interpretation and expand the scope of DR screening. We performed a retrospective review to compare Eyenuk's EyeArt software (Woodland Hills, CA) to Temple Ophthalmology optometry grading using the International Classification of Diabetic Retinopathy scale. Two hundred and sixty consecutive diabetic patients from the Temple Faculty Practice Internal Medicine clinic underwent 2-field retinal imaging. Classifications of the images by the software and optometrist were analyzed using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and McNemar's test. Ungradable images were analyzed to identify relationships with HbA1c, age, and ethnicity. Disagreements and a sample of 20% of agreements were adjudicated by a retina specialist. On patient level comparison, sensitivity for the software was 100%, while specificity was 77.78%. PPV was 19.15%, and NPV was 100%. The 38 disagreements between software and optometrist occurred when the optometrist classified a patient's images as non-referable while the software classified them as referable. Of these disagreements, a retina specialist agreed with the optometrist 57.9% the time (22/38). Of the agreements, the retina specialist agreed with both the program and the optometrist 96.7% of the time (28/29). There was a significant difference in numbers of ungradable photos in older patients (≥60) vs younger patients (<60) (p=0.003). The AI program showed high sensitivity with acceptable specificity for a screening algorithm. The high NPV indicates that the software is unlikely to miss DR but may refer patients unnecessarily.

Automated segmentation algorithm with deep learning framework for early detection of glaucoma

AbstractEarly stage of diagnosis of eye diseases through automatic analysis in the retinal image is the emerging technology in the area of retinopathy. Glaucoma is the primary reason for the loss of visibility in people around the world. The separation of the disc and the cup in the optic region is the technique used to identify glaucoma in the human retinal image. In this paper, superpixel segmentation, followed by Modified Kernel Fuzzy C‐Means (MKFCM) algorithm is used to segment the optic disc and optic cup. The proposed segmentation method achieves a maximum average of F‐score as 0.979, an average boundary distance as 10.016 pixels, and an average correlation coefficient of 0.949. To train convolutional neural networks (CNN), the segmented images obtained by the MKFCM segmentation algorithm is given as the input for the identification of glaucoma. This CNN uses the gray level co‐occurrence matrix features calculated from the segmented image. The experiment used for this study demonstrates that CNN gives superior categorization correctness and requires fewer figures of knowledge iterations than the original CNN. The accuracy obtained by this proposed method is 94.2%. The model will help to identify the proper class of severity of glaucoma in retinal images.

Toward optical coherence tomography on a chip: in vivo three-dimensional human retinal imaging using photonic integrated circuit-based arrayed waveguide gratings

In this work, we present a significant step toward in vivo ophthalmic optical coherence tomography and angiography on a photonic integrated chip. The diffraction gratings used in spectral-domain optical coherence tomography can be replaced by photonic integrated circuits comprising an arrayed waveguide grating. Two arrayed waveguide grating designs with 256 channels were tested, which enabled the first chip-based optical coherence tomography and angiography in vivo three-dimensional human retinal measurements. Design 1 supports a bandwidth of 22 nm, with which a sensitivity of up to 91 dB (830 µW) and an axial resolution of 10.7 µm was measured. Design 2 supports a bandwidth of 48 nm, with which a sensitivity of 90 dB (480 µW) and an axial resolution of 6.5 µm was measured. The silicon nitride-based integrated optical waveguides were fabricated with a fully CMOS-compatible process, which allows their monolithic co-integration on top of an optoelectronic silicon chip. As a benchmark for chip-based optical coherence tomography, tomograms generated by a commercially available clinical spectral-domain optical coherence tomography system were compared to those acquired with on-chip gratings. The similarities in the tomograms demonstrate the significant clinical potential for further integration of optical coherence tomography on a chip system.

DeepLab and Bias Field Correction Based Automatic Cone Photoreceptor Cell Identification with Adaptive Optics Scanning Laser Ophthalmoscope Images

The identification of cone photoreceptor cells is important for early diagnosing of eye diseases. We proposed automatic deep‐learning cone photoreceptor cell identification on adaptive optics scanning laser ophthalmoscope images. The proposed algorithm is based on DeepLab and bias field correction. Considering manual identification as reference, our algorithm is highly effective, achieving precision, recall, and F1 score of 96.7%, 94.6%, and 95.7%, respectively. To illustrate the performance of our algorithm, we present identification results for images with different cone photoreceptor cell distributions. The experimental results show that our algorithm can achieve accurate photoreceptor cell identification on images of human retinas, which is comparable to manual identification.

Optical Coherence Tomography for Ophthalmology Imaging.

Optical coherence tomography (OCT) is a depth-resolved imaging modality, which is able to achieve micrometer-scale resolution within biological tissue noninvasively. In the past 30years, researchers all around the world had made several essential efforts on techniques relevant to OCT. OCT has become a routine process for eye diseases with different types. In this chapter, the three important stages in the development of OCT are briefly illustrated, including the time domain OCT (TD-OCT), the frequency domain OCT (FD-OCT) and the optical coherence tomography angiography (OCTA). Each of the technique has made great progress for use on living human eye imaging in clinical applications. TD-OCT was first proposed and commercialized, which is able to achieve acceptable 2D depth-resolved cross-sectional images of human retina in vivo. FD-OCT was the upgraded OCT technique compared with TD-OCT. By capturing the coherent signal within the Fourier space, the FD-OCT could improve the image sensitivity compared with TD-OCT, and achieve dozens of kilo hertz imaging speed. OCTA is the newest developments of OCT technique, which is able to visualize the micro vasculature networks of human retina in vivo. With OCTA technique, the newest ophthalmologic OCT system is able to achieve detailed diagnosis for both micro-structure and vasculature abnormalities for clinical applications. The further development of OCT technique on imaging speed, contrast, resolution, field of view, and so on will make OCT to be a more powerful tool for clinical usages.

A vessel segmentation technique for retinal images

AbstractSegmentation of the human eye retinal image is an essential step for proper examination and diagnosis by the ophthalmologists or eye care specialists. A technique for vessel segmentation of retinal images is proposed. Retinal images are mostly low‐light images, which are first processed for enhancement of light as well as for detail amplification. Illumination of low‐light images is enhanced, and details are amplified using content‐adaptive filters. For extraction of vessels from retinal images, after low‐light and detail enhancement, the B‐cosfire filter is modified by including extraction of details and small elements, which may otherwise be ignored. A modified B‐cosfire filter is used to extract vessels while minimizing false edges and halo artifacts. The morphological opening is performed to crop vessels that are falsely segmented. The technique is contrasted with other existing methods in terms of accuracy using publicly available datasets. The proposed technique is tested on STARE, CHASE‐DB1, and DRIVE databases. The outcome of the proposed procedure has better accuracy, preserved edges, minimum noise, and artifacts than the state‐of‐the‐art techniques.

Reference variance and adjustment factor based wave atoms image speckle reduction method for optical coherence tomography images

Speckle noise in optical coherence tomography (OCT) images causes blurring of small targets and structural edges in the image, which reduces the quality of OCT images. To address this problem, a reference variance and adjustment factor based wave atoms image speckle reduction method for OCT images is proposed. First, reference image and adjustment factor were introduced, and reference image speckle noise reference variance was adjusted using adjustment factor to obtain the target image speckle noise variance. Then, a wave atoms threshold equation including the reference variance and the adjustment factor was constructed to obtain an accurate wave atoms noise reduction threshold. On this basis, speckle noise was shrunk using the wave atoms noise reduction threshold to achieve speckle noise reduction in OCT images, and OCT images with clear small targets and sharp structure edges were obtained. The effectiveness and advantages of the proposed method were illustrated with human eye OCT image and retinal OCT image.

Visible light optical coherence tomography angiography (vis-OCTA) facilitates local microvascular oximetry in the human retina.

We report herein the first visible light optical coherence tomography angiography (vis-OCTA) for human retinal imaging. Compared to the existing vis-OCT systems, we devised a spectrometer with a narrower bandwidth to increase the spectral power density for OCTA imaging, while retaining the major spectral contrast in the blood. We achieved a 100 kHz A-line rate, the fastest acquisition speed reported so far for human retinal vis-OCT. We rigorously optimized the imaging protocol such that a single acquisition took < 6 seconds with a field of view (FOV) of 3×7.8 mm2. The angiography enables accurate localization of microvasculature down to the capillary level and thus enables oximetry at vessels < 100 µm in diameter. We demonstrated microvascular hemoglobin oxygen saturation (sO2) at the feeding and draining vessels at the perifoveal region. The longitudinal repeatability was assessed by < 5% coefficient of variation (CV). The unique capabilities of our vis-OCTA system may allow studies on the role of microvascular oxygen in various retinal pathologies.

Texture preservation and speckle reduction in poor optical coherence tomography using the convolutional neural network.

For a poor quality optical coherence tomography (OCT) image, quality enhancement is limited to speckle residue and edge blur as well as texture loss, especially at the background region near edges. To solve this problem, in this paper we propose a de-speckling method based on the convolutional neural network (CNN). In the proposed method, we use a deep nonlinear CNN mapping model in the serial architecture, here named as OCTNet. Our OCTNet in the proposed method can fully utilize the deep information on speckles and edges as well as fine textures of an original OCT image. And also we construct an available pertinent dataset by combining three existing methods to train the model. With the proposed method, we can accurately get the speckle noise from an original OCT image. We test our method on four experimental human retinal OCT images and also compare it with three state-of-the-art methods, including the adaptive complex diffusion (ACD) method and the curvelet shrinkage (Curvelet) method as well as the shearlet-based total variation (STV) method. The performance of these methods is quantitatively evaluated in terms of image distinguishability, contrast, smoothness and edge sharpness, and also qualitatively analyzed at aspects of speckle reduction, texture protection and edge preservation. The experimental results show that our OCTNet can reduce the speckle noise and protect the structural information as well as preserve the edge features effectively and simultaneously, even where the background region near edges. And also our OCTNet has full advantages on excellent generalization, adaptiveness, robust and batch performance. These advantages make our method be suitable to process a great mass of different images rapidly without any parameter fine-turning under a time-constrained real-time situation.

Is oblique scanning laser ophthalmoscope applicable to human ocular optics? A feasibility study using an eye model for volumetric imaging.

Oblique scanning laser ophthalmoscopy (oSLO) is a novel imaging modality to provide volumetric retinal imaging without depth sectioning over a large field of view (FOV). It has been successfully demonstrated in vivo in rodent eyes for volumetric fluorescein angiography (vFA). However, engineering oSLO for human retinal imaging is challenging because of the low numerical aperture (NA) of human ocular optics. To overcome this challenge, we implement optical designs to (a) increase the angle of the intermediate image under Scheimpflug condition, and (b) expand the magnification in the depth dimension with cylindrical lens to enable sufficient sampling density. In addition, we adopt a scanning-and-descaning strategy, resulting in a compact oSLO system. We experimentally show that the current setup can achieve a FOV of ~3 × 6 × 0.8 mm3 , and the transverse and axial resolutions of 7 and 41 μm, respectively. This feasibility study serves an important step for future in vivo human retinal imaging.

High-resolution in-vivo human retinal imaging using full-field OCT with optical stabilization of axial motion.

Time-domain full-field OCT (FF-OCT) represents an imaging modality capable of recording high-speed en-face sections of a sample at a given depth. One of the biggest challenges to transfer this technique to image in-vivo human retina is the presence of continuous involuntary head and eye axial motion during image acquisition. In this paper, we demonstrate a solution to this problem by implementing an optical stabilization in an FF-OCT system. This was made possible by combining an FF-OCT system, an SD-OCT system, and a high-speed voice-coil translation stage. B-scans generated by the SD-OCT were used to measure the retina axial position and to drive the position of the high-speed voice coil translation stage, where the FF-OCT reference arm is mounted. Closed-loop optical stabilization reduced the RMS error by a factor of 7, significantly increasing the FF-OCT image acquisition efficiency. By these means, we demonstrate the capacity of the FF-OCT to resolve cone mosaic as close as 1.5 o from the fovea center with high consistency and without using adaptive optics.

Mathematical Morphology and Optimum Principal Curvature Based Segmentation of Blood Vessels in Human Retinal Fundus Images

The retinal abnormalities and diagnosis of Diabetic Retinopathy (DR), Glaucoma are accomplished by extraction of vessel network in human retinal images. An accurate segmentation is required for the pathological analysis. Various researchers proposed many automated systems for vessel segmentation, still this process needs an improvement due to the presence of abnormalities, different magnitude, dimension of the vessels, non-uniform lighting and variable structure of the retina. The proposed work is a new method for retinal vessel segmentation, which consists of three phases, (i) The vessels network is enhanced by using Contrast Limited Adaptive Histogram Equalization(CLAHE) and Median filtering techniques (ii) the smoothened image is segmented based on mathematical morphology and maximum principal curvature followed by cleaning operation to remove the small objects, (iii) the segmented image is compared with hand labeled Ground Truth image and is evaluated with the True Positive, False Positive , True Negative and False Negative parameters. The performance of this work is tested with the images existing in DRIVE database. This work achieves 0.965 Accuracy, 0.752 Sensitivity and 0.989 Specificity.