A foundation model for generalizable disease detection from retinal images.

Medical artificial intelligence (AI) offers great potential for recognizing signs of health conditions in retinal images and expediting the diagnosis of eye diseases and systemic disorders1. However, the development of AI models requires substantial annotation and models are usually task-specific with limited generalizability to different clinical applications2. Here, we present RETFound, a foundation model for retinal images that learns generalizable representations from unlabelled retinal images and provides a basis for label-efficient model adaptation in several applications. Specifically, RETFound is trained on 1.6 million unlabelled retinal images by means of self-supervised learning and then adapted to disease detection tasks with explicit labels. We show that adapted RETFound consistently outperforms several comparison models in the diagnosis and prognosis of sight-threatening eye diseases, as well as incident prediction of complex systemic disorders such as heart failure and myocardial infarction with fewer labelled data. RETFound provides a generalizable solution to improve model performance and alleviate the annotation workload of experts to enable broad clinical AI applications from retinal imaging.

Ultrasound in ocular oncology: Technical advances, clinical applications, and limitations.

Due to its accessibility and ability for real-time image acquisition of ocular structures, ultrasound has high utility in the visualization of the eye, especially in ocular oncology. In this minireview, we summarize the technical rationale and applications of ultrasound modalities, A-scan, B-scan, high-frequency ultrasound biomicroscopy (UBM), and Doppler measurement. A-scan ultrasound uses a transducer of 7-11 MHz, making it useful for determining the echogenicity of ocular tumors (7-8 MHz) and measuring the axial length of the eye (10-11 MHz). B-scan ultrasound operates at 10-20 MHz, which can be used for measuring posterior ocular tumors while UBM operates at 40-100 MHz to evaluate anterior ocular structures. Doppler ultrasonography allows for the detection of tumor vascularization. While ultrasonography has numerous clinical applications due to its favorable penetration compared with optical coherence tomography, it is still limited by its relatively lower resolution. Ultrasound also requires an experienced sonographer due to the need for accurate probe localization to areas of interest.

Deep learning in optical coherence tomography: Where are the gaps?

Optical coherence tomography (OCT) is a non-invasive optical imaging modality, which provides rapid, high-resolution and cross-sectional morphology of macular area and optic nerve head for diagnosis and managing of different eye diseases. However, interpreting OCT images requires experts in both OCT images and eye diseases since many factors such as artefacts and concomitant diseases can affect the accuracy of quantitative measurements made by post-processing algorithms. Currently, there is a growing interest in applying deep learning (DL) methods to analyse OCT images automatically. This review summarises the trends in DL-based OCT image analysis in ophthalmology, discusses the current gaps, and provides potential research directions. DL in OCT analysis shows promising performance in several tasks: (1) layers and features segmentation and quantification; (2) disease classification; (3) disease progression and prognosis; and (4) referral triage level prediction. Different studies and trends in the development of DL-based OCT image analysis are described and the following challenges are identified and described: (1) public OCT data are scarce and scattered; (2) models show performance discrepancies in real-world settings; (3) models lack of transparency; (4) there is a lack of societal acceptance and regulatory standards; and (5) OCT is still not widely available in underprivileged areas. More work is needed to tackle the challenges and gaps, before DL is further applied in OCT image analysis for clinical use.

THE WAY PATIENTS SEE FLOATERS: Widefield Dynamic Scanning Laser Ophthalmoscopy Imaging of Vitreous Abnormalities.

PURPOSE: To investigate the use of dynamic widefield scanning laser ophthalmoscopy (SLO) and B-scan ultrasonography in imaging vitreous abnormalities in patients with complaints of floaters. METHODS: Twenty-one patients underwent both dynamic SLO and B-scan ultrasonography to image their vitreous abnormalities. After reviewing these videos, patients graded each imaging technique on a scale of 1 to 10, based on how closely it represented their visual perception of floaters. RESULTS: The mean age of the patients (12 women and nine men) was 47.7 ± 18.5 years. The patients graded a median score of nine for SLO imaging (mean = 8.43) compared with a median score of 5 (mean = 4.95) for ultrasound ( P = 0.001). Widefield SLO imaging demonstrated three-dimensional interconnectivity within the condensations of the formed vitreous that exhibited translational and rotational movements with eye saccades. CONCLUSION: Floaters are a common complaint, but it is difficult to know whether imaging findings of the vitreous correlate to what patients perceive. Widefield SLO seems to image vitreous abnormalities related to how patients perceive their own floaters better than B-scan ultrasonography. Despite the term "floaters", the vitreous abnormalities in the videos seemed to be manifestations of a complex three-dimensional degeneration of the vitreous framework.

[Chinese expert consensus on the clinical visual electrophysiologic examination terminology (2023)].

Visual electrophysiology is an objective examination method for assessing visual function. As one of the important ophthalmic clinical examinations, it is widely used in the diagnosis, differential diagnosis, follow-up and visual function identification of diseases. Based on a number of standards and guidelines published by the International Society of Clinical Visual Electrophysiology in recent years, in combination with the recent clinical practice and research progress in China, the experts in the Visual Physiology Group of Ophthalmology Branch of Chinese Medical Association and Visual Physiology Group of Chinese Ophthalmologist Association have formed consensus opinions to help Chinese ophthalmologists standardize the use of clinical visual electrophysiologic terminology and to promote the further standardization of clinical visual electrophysiologic examination in China.

Portable Anterior Eye Segment Imaging System for Teleophthalmology.

Objective: This study aims to compare a new prototype for a portable anterior eye segment imaging system with the standard method for ophthalmology examination. Methods: The new imaging system consisted of two IMX219 Arducam autofocus sensors (Arducam, China, Nanjing) for Raspberry Pi V2 camera module connected to a Raspberry Pi Zero W (Raspberry Pi Foundation, UK, Cambridge) that clips to a wearable headset. The 2D videos of the anterior eye segment were recorded with the new system and a 720p FaceTime HD camera (Apple, Cupertino, CA). Afterward, ophthalmologists evaluated the videos using a standard clinical eye examination form. These evaluations were compared with the standard slit-lamp clinical assessment performed during the patient's visit. Results: Thirty-five eyes were evaluated. The sensitivity and specificity percentages were statistically significant between the two imaging modalities (P ≤ 0.001). The evaluations performed from videos obtained with the new imaging system had better sensitivity and specificity percentages overall. However, statistically significant differences were only observed in cornea, anterior chamber, iris, and lens. Conclusions: Specificity percentages were higher than sensitivity percentages in both imaging modalities, indicating that video evaluations are less accurate for pathological screening. Nevertheless, the new system evaluations were significantly better than the webcam evaluations. Translational Relevance: This study presented an alternative system to assess eye conditions for telemedicine, one that provides more details than the current standard and uses new wearable headsets technologies.

Deep learning for ultra-widefield imaging: a scoping review.

PURPOSE: This article is a scoping review of published and peer-reviewed articles using deep-learning (DL) applied to ultra-widefield (UWF) imaging. This study provides an overview of the published uses of DL and UWF imaging for the detection of ophthalmic and systemic diseases, generative image synthesis, quality assessment of images, and segmentation and localization of ophthalmic image features. METHODS: A literature search was performed up to August 31st, 2021 using PubMed, Embase, Cochrane Library, and Google Scholar. The inclusion criteria were as follows: (1) deep learning, (2) ultra-widefield imaging. The exclusion criteria were as follows: (1) articles published in any language other than English, (2) articles not peer-reviewed (usually preprints), (3) no full-text availability, (4) articles using machine learning algorithms other than deep learning. No study design was excluded from consideration. RESULTS: A total of 36 studies were included. Twenty-three studies discussed ophthalmic disease detection and classification, 5 discussed segmentation and localization of ultra-widefield images (UWFIs), 3 discussed generative image synthesis, 3 discussed ophthalmic image quality assessment, and 2 discussed detecting systemic diseases via UWF imaging. CONCLUSION: The application of DL to UWF imaging has demonstrated significant effectiveness in the diagnosis and detection of ophthalmic diseases including diabetic retinopathy, retinal detachment, and glaucoma. DL has also been applied in the generation of synthetic ophthalmic images. This scoping review highlights and discusses the current uses of DL with UWF imaging, and the future of DL applications in this field.

Preclinical ocular microvascular changes in juvenile dermatomyositis: A pilot optical coherence tomography angiography study.

OBJECTIVES: To evaluate the macular and optic nerve head (ONH) vascular density, foveal avascular zone area, and outer retina and choriocapillaris flow in juvenile dermatomyositis (JDM) using optical coherence tomography angiography (OCTA). METHODS: Ten eyes of 10 patients with JDM and 15 age and sex-matched healthy controls were investigated in this prospective, cross-sectional study. The superficial capillary plexus (SCP) and deep capillary plexus (DCP), ONH, foveal avascular zone (FAZ) parameters, the flow area of the outer retina, and choriocapillaris were evaluated using OCTA. RESULTS: Vessel density (VD) of the parafovea (p = 0.036) and parafoveal subregions (p = 0.041 for superior hemifield, p = 0.031 for inferior hemifield, p = 0.012 for superior, p = 0.019 for nasal, p = 0.026 for inferior, and p = 0.048 for temporal) in DCP were significantly lower in the JDM group compared to healthy controls. A high inverse correlation between disease duration and these parameters was found except parafoveal superior VD in DCP. There was no significant difference between the groups in VD parameters of SCP and ONH, FAZ parameters, outer retina, and choriocapillaris flow area as well as thickness parameters. (p > 0.05 for all). Furthermore, ROC analysis revealed that all parafoveal DCP parameters showed good ability to differentiate JDM from healthy controls. CONCLUSIONS: We demonstrated a decreased vessel density in the deep parafoveal region in JDM. As a result, we hypothesized that OCTA could detect retinal microvascular changes in JDM patients who did not have clinical evidence of ocular involvement.

Development and validation of an offline deep learning algorithm to detect vitreoretinal abnormalities on ocular ultrasound.

Purpose: We describe our offline deep learning algorithm (DLA) and validation of its diagnostic ability to identify vitreoretinal abnormalities (VRA) on ocular ultrasound (OUS). Methods: Enrolled participants underwent OUS. All images were classified as normal or abnormal by two masked vitreoretinal specialists (AS, AM). A data set of 4902 OUS images was collected, and 4740 images of satisfactory quality were used. Of this, 4319 were processed for further training and development of DLA, and 421 images were graded by vitreoretinal specialists (AS and AM) to obtain ground truth. The main outcome measures were sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and area under receiver operating characteristic (AUROC). Results: Our algorithm demonstrated high sensitivity and specificity in identifying VRA on OUS ([90.8%; 95% confidence interval (CI): 86.1-94.3%] and [97.1% (95% CI: 93.7-98.9%], respectively). PPV and NPV of the algorithm were also high ([97.0%; 95% CI: 93.7-98.9%] and [90.8%; 95% CI: 86.2-94.3%], respectively). The AUROC was high at 0.939, and the intergrader agreement was nearly perfect with Cohen's kappa of 0.938. The model demonstrated high sensitivity in predicting vitreous hemorrhage (100%), retinal detachment (97.4%), and choroidal detachment (100%). Conclusion: Our offline DLA software demonstrated reliable performance (high sensitivity, specificity, AUROC, PPV, NPV, and intergrader agreement) for predicting VRA on OUS. This might serve as an important tool for the ophthalmic technicians who are involved in community eye screening at rural settings where trained ophthalmologists are not available.

Foundational Considerations for Artificial Intelligence Using Ophthalmic Images.

IMPORTANCE: The development of artificial intelligence (AI) and other machine diagnostic systems, also known as software as a medical device, and its recent introduction into clinical practice requires a deeply rooted foundation in bioethics for consideration by regulatory agencies and other stakeholders around the globe. OBJECTIVES: To initiate a dialogue on the issues to consider when developing a bioethically sound foundation for AI in medicine, based on images of eye structures, for discussion with all stakeholders. EVIDENCE REVIEW: The scope of the issues and summaries of the discussions under consideration by the Foundational Principles of Ophthalmic Imaging and Algorithmic Interpretation Working Group, as first presented during the Collaborative Community on Ophthalmic Imaging inaugural meeting on September 7, 2020, and afterward in the working group. FINDINGS: Artificial intelligence has the potential to improve health care access and patient outcome fundamentally while decreasing disparities, lowering cost, and enhancing the care team. Nevertheless, substantial concerns exist. Bioethicists, AI algorithm experts, as well as the Food and Drug Administration and other regulatory agencies, industry, patient advocacy groups, clinicians and their professional societies, other provider groups, and payors (i.e., stakeholders) working together in collaborative communities to resolve the fundamental ethical issues of nonmaleficence, autonomy, and equity are essential to attain this potential. Resolution impacts all levels of the design, validation, and implementation of AI in medicine. Design, validation, and implementation of AI warrant meticulous attention. CONCLUSIONS AND RELEVANCE: The development of a bioethically sound foundation may be possible if it is based in the fundamental ethical principles of nonmaleficence, autonomy, and equity for considerations for the design, validation, and implementation for AI systems. Achieving such a foundation will be helpful for continuing successful introduction into medicine before consideration by regulatory agencies. Important improvements in accessibility and quality of health care, decrease in health disparities, and lower cost thereby can be achieved. These considerations should be discussed with all stakeholders and expanded on as a useful initiation of this dialogue.

Automatic Recognition of Ocular Surface Diseases on Smartphone Images Using Densely Connected Convolutional Networks.

Ocular surface disorder is one of common and prevalence eye diseases and complex to be recognized accurately. This work presents automatic classification of ocular surface disorders in accordance with densely connected convolutional networks and smartphone imaging. We use various smartphone cameras to collect clinical images that contain normal and abnormal, and modify end-to-end densely connected convolutional networks that use a hybrid unit to learn more diverse features, significantly reducing the network depth, the total number of parameters and the float calculation. The validation results demonstrate that our proposed method provides a promising and effective strategy to accurately screen ocular surface disorders. In particular, our deeply learned smartphone photographs based classification method achieved an average automatic recognition accuracy of 90.6%, while it is conveniently used by patients and integrated into smartphone applications for automatic patient-self screening ocular surface diseases without seeing a doctor in person in a hospital.

Multi-task Learning Based Ocular Disease Discrimination and FAZ Segmentation Utilizing OCTA Images.

In this paper, we proposed and validated a multi-task based deep learning method for simultaneously segmenting the foveal avascular zone (FAZ) and classifying three ocular disease related states (normal, diabetic, and myopia) utilizing optical coherence tomography angiography (OCTA) images. The essential motivation of this work is that reliable predictions on disease states may be made based on features extracted from a segmentation network, by sharing a same encoder between the classification network and the segmentation network. In this study, a cotraining network structure was designed for simultaneous ocular disease discrimination and FAZ segmentation. Specifically, we made use of a classification head following a segmentation network's encoder, so that the classification branch used the feature information extracted in the segmentation branch to improve the classification results. The performance of our proposed network structure has been tested and validated on the FAZID dataset, with the best Dice and Jaccard being 0.9031±0.0772 and 0.8302 ±0.0990 for FAZ segmentation, and the best Accuracy and Kappa being 0.7533 and 0.6282 for classifying three ocular disease related states.Clinical Relevance- This work provides a useful tool for segmenting FAZ and discriminating three ocular disease related states utilizing OCTA images, which has a great clinical potential in ocular disease screening and biomarker delivering.

Ocular Ultrasound Abnormalities and Optic Nerve Sheath Diameter in Dogs and Cats.

Indications for, technique, and findings for normal and abnormal ocular ultrasound are discussed, with specific sonographic findings, images, differential diagnoses, and other considerations. Because the eye is a fluid-filled structure, ultrasound can be used as a screening test when pathology prevents direct examination. Structural abnormalities, such as lens luxation, retinal detachments, and intraocular and orbital masses, also may be defined better using point-of-care ultrasound. Details on additional ophthalmic diagnostics, treatment, and prognosis are not covered.