Search

PURPOSE: The aim of this study is to investigate the efficacy of a mobile platform that combines smartphone-based retinal imaging with automated grading for determining the presence of referral-warranted diabetic retinopathy (RWDR). METHODS: A smartphone-based camera (RetinaScope) was used by non-ophthalmic personnel to image the retina of patients with diabetes. Images were analyzed with the Eyenuk EyeArt® system, which generated referral recommendations based on presence of diabetic retinopathy (DR) and/or markers for clinically significant macular oedema. Images were independently evaluated by two masked readers and categorized as refer/no refer. The accuracies of the graders and automated interpretation were determined by comparing results to gold standard clinical diagnoses. RESULTS: A total of 119 eyes from 69 patients were included. RWDR was present in 88 eyes (73.9%) and in 54 patients (78.3%). At the patient-level, automated interpretation had a sensitivity of 87.0% and specificity of 78.6%; grader 1 had a sensitivity of 96.3% and specificity of 42.9%; grader 2 had a sensitivity of 92.5% and specificity of 50.0%. At the eye-level, automated interpretation had a sensitivity of 77.8% and specificity of 71.5%; grader 1 had a sensitivity of 94.0% and specificity of 52.2%; grader 2 had a sensitivity of 89.5% and specificity of 66.9%. DISCUSSION: Retinal photography with RetinaScope combined with automated interpretation by EyeArt achieved a lower sensitivity but higher specificity than trained expert graders. Feasibility testing was performed using non-ophthalmic personnel in a retina clinic with high disease burden. Additional studies are needed to assess efficacy of screening diabetic patients from general population.

To compare general ophthalmologists, retina specialists, and the EyeArt Artificial Intelligence (AI) system to the clinical reference standard for detecting more than mild diabetic retinopathy (mtmDR).Prospective, pivotal, multicenter trial conducted from April 2017 to May 2018.Participants were aged ≥ 18 years who had diabetes mellitus and underwent dilated ophthalmoscopy. A total of 521 of 893 participants met these criteria and completed the study protocol.Participants underwent 2-field fundus photography (macula centered, disc centered) for the EyeArt system, dilated ophthalmoscopy, and 4-widefield stereoscopic dilated fundus photography for reference standard grading.For mtmDR detection, sensitivity and specificity of EyeArt gradings of 2-field, fundus photographs and ophthalmoscopy grading versus a rigorous clinical reference standard comprising Reading Center grading of 4-widefield stereoscopic dilated fundus photographs using the ETDRS severity scale. The AI system provided automatic eye-level results regarding mtmDR.Overall, 521 participants (999 eyes) at 10 centers underwent dilated ophthalmoscopy: 406 by nonretina and 115 by retina specialists. Reading Center graded 207 positive and 792 eyes negative for mtmDR. Of these 999 eyes, 26 eyes were ungradable by the EyeArt system, leaving 973 eyes with both EyeArt and Reading Center gradings. Retina specialists correctly identified 22 of 37 eyes as positive (sensitivity 59.5%) and 182 of 184 eyes as negative (specificity 98.9%) for mtmDR versus the EyeArt AI system that identified 36 of 37 as positive (sensitivity 97%) and 162 of 184 eyes as negative (specificity of 88%) for mtmDR. General ophthalmologists correctly identified 35 of 170 eyes as positive (sensitivity 20.6%) and 607 of 608 eyes as negative (specificity 99.8%) for mtmDR compared with the EyeArt AI system that identified 164 of 170 as positive (sensitivity 96.5%) and 525 of 608 eyes as negative (specificity 86%) for mtmDR.The AI system had a higher sensitivity for detecting mtmDR than either general ophthalmologists or retina specialists compared with the clinical reference standard. It can potentially serve as a low-cost point-of-care diabetic retinopathy detection tool and help address the diabetic eye screening burden.

Background: Current manual diabetic retinopathy (DR) screening using eye care experts cannot scale to screen the growing population of diabetes patients who are at risk for vision loss. EyeArt system is an automated, cloud-based artificial intelligence (AI) eye screening technology designed to easily detect referral-warranted DR immediately through automated analysis of patient's retinal images. Methods: This retrospective study assessed the diagnostic efficacy of the EyeArt system v2.0 analyzing 850,908 fundus images from 101,710 consecutive patient visits, collected from 404 primary care clinics. Presence or absence of referral-warranted DR (more than mild nonproliferative DR [NPDR]) was automatically detected by the EyeArt system for each patient encounter, and its performance was compared against a clinical reference standard of quality-assured grading by rigorously trained certified ophthalmologists and optometrists. Results: Of the 101,710 visits, 75.7% were nonreferable, 19.3% were referable to an eye care specialist, and in 5.0%, the DR level was unknown as per the clinical reference standard. EyeArt screening had 91.3% (95% confidence interval [CI]: 90.9–91.7) sensitivity and 91.1% (95% CI: 90.9–91.3) specificity. For 5446 encounters with potentially treatable DR (more than moderate NPDR and/or diabetic macular edema), the system provided a positive “refer” output to 5363 encounters achieving sensitivity of 98.5%. Conclusions: This study captures variations in real-world clinical practice and shows that an AI DR screening system can be safe and effective in the real world. This study demonstrates the value of this easy-to-use, automated tool for endocrinologists, diabetologists, and general practitioners to address the growing need for DR screening and monitoring.

Purpose Retinal screening examinations can prevent vision loss resulting from diabetes but are costly and highly underused. We hypothesized that artificial intelligence-assisted nonmydriatic point-of-care screening administered during primary care visits would increase the adherence to recommendations for follow-up eye care in patients with diabetes. Design Prospective cohort study. Participants Adults 18 years of age or older with a clinical diagnosis of diabetes being cared for in a metropolitan primary care practice for low-income patients. Methods All participants underwent nonmydriatic fundus photography followed by automated retinal image analysis with human supervision. Patients with positive or inconclusive screening results were referred for comprehensive ophthalmic evaluation. Adherence to referral recommendations was recorded and compared with the historical adherence rate from the same clinic. Main Outcome Measure Rate of adherence to eye screening recommendations. Results By automated screening, 8.3% of the 180 study participants had referable diabetic eye disease, 13.3% had vision-threatening disease, and 29.4% showed inconclusive results. The remaining 48.9% showed negative screening results, confirmed by human overread, and were not referred for follow-up ophthalmic evaluation. Overall, the automated platform showed a sensitivity of 100% (confidence interval, 92.3%–100%) in detecting an abnormal screening results, whereas its specificity was 65.7% (confidence interval, 57.0%–73.7%). Among patients referred for follow-up ophthalmic evaluation, the adherence rate was 55.4% at 1 year compared with the historical adherence rate of 18.7% (P Conclusions Implementation of an automated diabetic retinopathy screening system in a primary care clinic serving a low-income metropolitan patient population improved adherence to follow-up eye care recommendations while reducing referrals for patients with low-risk features.

Importance Diabetic retinopathy (DR) is a leading cause of blindness in adults worldwide. Early detection and intervention can prevent blindness; however, many patients do not receive their recommended annual diabetic eye examinations, primarily owing to limited access. Objective To evaluate the safety and accuracy of an artificial intelligence (AI) system (the EyeArt Automated DR Detection System, version 2.1.0) in detecting both more-than-mild diabetic retinopathy (mtmDR) and vision-threatening diabetic retinopathy (vtDR). Design, Setting, and Participants A prospective multicenter cross-sectional diagnostic study was preregistered (NCT03112005) and conducted from April 17, 2017, to May 30, 2018. A total of 942 individuals aged 18 years or older who had diabetes gave consent to participate at 15 primary care and eye care facilities. Data analysis was performed from February 14 to July 10, 2019. Interventions Retinal imaging for the autonomous AI system and Early Treatment Diabetic Retinopathy Study (ETDRS) reference standard determination. Main Outcomes and Measures Primary outcome measures included the sensitivity and specificity of the AI system in identifying participants’ eyes with mtmDR and/or vtDR by 2-field undilated fundus photography vs a rigorous clinical reference standard comprising reading center grading of 4 wide-field dilated images using the ETDRS severity scale. Secondary outcome measures included the evaluation of imageability, dilated-if-needed analysis, enrichment correction analysis, worst-case imputation, and safety outcomes. Results Of 942 consenting individuals, 893 patients (1786 eyes) met the inclusion criteria and completed the study protocol. The population included 449 men (50.3%). Mean (SD) participant age was 53.9 (15.2) years (median, 56; range, 18-88 years), 655 were White (73.3%), and 206 had type 1 diabetes (23.1%). Sensitivity and specificity of the AI system were high in detecting mtmDR (sensitivity: 95.5%; 95% CI, 92.4%-98.5% and specificity: 85.0%; 95% CI, 82.6%-87.4%) and vtDR (sensitivity: 95.1%; 95% CI, 90.1%-100% and specificity: 89.0%; 95% CI, 87.0%-91.1%) without dilation. Imageability was high without dilation, with the AI system able to grade 87.4% (95% CI, 85.2%-89.6%) of the eyes with reading center grades. When eyes with ungradable results were dilated per the protocol, the imageability improved to 97.4% (95% CI, 96.4%-98.5%), with the sensitivity and specificity being similar. After correcting for enrichment, the mtmDR specificity increased to 87.8% (95% CI, 86.3%-89.5%) and the sensitivity remained similar; for vtDR, both sensitivity (97.0%; 95% CI, 91.2%-100%) and specificity (90.1%; 95% CI, 89.4%-91.5%) improved. Conclusions and Relevance This prospective multicenter cross-sectional diagnostic study noted safety and accuracy with use of the EyeArt Automated DR Detection System in detecting both mtmDR and, for the first time, vtDR, without physician assistance. These findings suggest that improved access to accurate, reliable diabetic eye examinations may increase adherence to recommended annual screenings and allow for accelerated referral of patients identified as having vtDR.

We have developed a novel technique to explain results of an AI system and demonstrate its ability in explaining the EyeArt AI eye screening system for DR. Besides providing clinically consistent and accurate results, AI systems in healthcare must be explainable (be able to provide reasons for their decisions). This is paramount in making them trustworthy. The EyeArt system is intended for use by healthcare providers to screen for referable DR using color fundus photographs from 18 years or older diabetic patients who do not have persistent visual impairment. It is designed to provide screening results in under a minute and enable DR screening at point of care in endocrinology and primary care clinics. Figure shows that the explainability technique highlights image regions that most significantly contribute to the EyeArt system’s results. These highlighted regions correspond to lesions marked by an expert human grader. In a previous study on 107,001 consecutive patient visits, the EyeArt system has been shown to achieve sensitivity of 91.3% (95% CI: 90.9%-91.7%), specificity of 91.1% (95% CI: 90.9%-91.3%), NPV of 97.6% (95% CI: 97.5%-97.7%), and PPV of 72.5% (95% CI: 71.9%-73.0%). The work presented here provides evidence that the EyeArt system is explainable and has been designed to report DR screening results based on lesions considered important by the ophthalmology community for DR severity grading. Disclosure N. Parekh: Employee; Self; Eyenuk Inc. M. Bhaskaranand: Employee; Self; Eyenuk Inc. C. Ramachandra: Employee; Self; Eyenuk Inc. S. Bhat: Employee; Self; Eyenuk Inc. K. Solanki: Employee; Self; Eyenuk Inc. Stock/Shareholder; Self; Eyenuk Inc. Funding National Institutes of Health

Purpose: Diabetic retinopathy (DR) is a leading cause of blindness, yet a majority of adults with diabetes are noncompliant with annual ophthalmic screening recommendations. We hypothesized that point-of-care automated DR could increase the rate of compliance with recommended retinal surveillance in patients with diabetes when administered during primary care visits. Methods: We performed a prospective cohort study of adults ages 18 or older with diabetes seen at a primary care clinic from 3/1/18 to 8/31/18. A retrospective chart review of adults with diabetes from the same clinic was also conducted and served as a historical control. Participants were screened using non-mydriatic fundus photography and automated DR screening software (Eyenuk Inc.). Those with positive or inconclusive screening results were referred to ophthalmology for comprehensive evaluation. Rates of compliance with referral recommendations were compared between the study participants and historical control. Results: One hundred and sixty-five adults with diabetes were included in the prospective study cohort. Twelve (7.3%) screened positive for referable DR, 23 (13.9%) screened positive for vision threatening DR, and 47 (28.4%) had an inconclusive screening result. Those that screened negative (83 participants) were all confirmed not to have DR by independent review from a retina specialist. The rate of compliance with follow-up eye care in referred participants was 54.9% compared to a historical rate of 18.7% (p Conclusions: The use of automated DR screening at our institution's primary care clinic reduced referrals for those with no retinopathy and increased the rate of compliance with follow-up eye care in those with positive or inconclusive screening results. These findings suggest that implementation of automated screening could effectively increase compliance with ophthalmic care among those with referable DR. Disclosure J.C. Liu: None. S. Ramchal: None. E. Gibson: None. K. Solanki: Employee; Self; Eyenuk Inc. Stock/Shareholder; Self; Eyenuk Inc. M. Bhaskaranand: Employee; Self; Eyenuk Inc. E. Fondahn: Consultant; Self; Greeley Company. R. Rajagopal: None.

Diabetic retinopathy (DR)-a common complication of diabetes-is the leading cause of vision loss among the working-age population in the western world. DR is largely asymptomatic, but if detected at early stages the progression to vision loss can be significantly slowed. With the increasing diabetic population there is an urgent need for automated DR screening and monitoring. To address this growing need, in this article we discuss an automated DR screening tool and extend it for automated estimation of microaneurysm (MA) turnover, a potential biomarker for DR risk.The DR screening tool automatically analyzes color retinal fundus images from a patient encounter for the various DR pathologies and collates the information from all the images belonging to a patient encounter to generate a patient-level screening recommendation. The MA turnover estimation tool aligns retinal images from multiple encounters of a patient, localizes MAs, and performs MA dynamics analysis to evaluate new, persistent, and disappeared lesion maps and estimate MA turnover rates.The DR screening tool achieves 90% sensitivity at 63.2% specificity on a data set of 40 542 images from 5084 patient encounters obtained from the EyePACS telescreening system. On a subset of 7 longitudinal pairs the MA turnover estimation tool identifies new and disappeared MAs with 100% sensitivity and average false positives of 0.43 and 1.6 respectively.The presented automated tools have the potential to address the growing need for DR screening and monitoring, thereby saving vision of millions of diabetic patients worldwide.

We design a video encoding scheme that is suited for applications such as unmanned aerial vehicle (UAV) video surveillance where the encoder complexity needs to be low. Our low complexity encoder predicts frames using the global motion information available in UAVs and thus achieves lower complexity and more than 40% BD-rate savings for fly-over videos compared to a complexity-constrained H.264 encoder with motion estimation restricted to 8 $\times$ 8 blocks and half pixel accuracy. We also incorporate a spectral entropy based bit allocation scheme into this encoder to achieve near constant quality within groups of pictures (GOPs) at the cost of small increases in delay and complexity, and a small drop in compression efficiency. Both these encoders with their corresponding low complexity “matched” decoders provide significant gains of more than 49% BD-rate savings over the Wyner-Ziv based DISCOVER codec which has a low complexity encoder and a high complexity decoder. Furthermore, for videos where the global motion is spatially consistent within 2 $\times$ 2 blocks, we show that the computational complexity of these proposed encoders can be significantly reduced with only about 1% BD-rate increase.

Purpose We examined the sensitivity and specificity of an automated algorithm for detecting referral-warranted diabetic retinopathy (DR) on Optos ultrawidefield (UWF) pseudocolour images. Methods Patients with diabetes were recruited for UWF imaging. A total of 383 subjects (754 eyes) were enrolled. Nonproliferative DR graded to be moderate or higher on the 5-level International Clinical Diabetic Retinopathy (ICDR) severity scale was considered as grounds for referral. The software automatically detected DR lesions using the previously trained classifiers and classified each image in the test set as referral-warranted or not warranted. Sensitivity, specificity and the area under the receiver operating curve (AUROC) of the algorithm were computed. Results The automated algorithm achieved a 91.7%/90.3% sensitivity (95% CI 90.1–93.9/80.4–89.4) with a 50.0%/53.6% specificity (95% CI 31.7–72.8/36.5–71.4) for detecting referral-warranted retinopathy at the patient/eye levels, respectively; the AUROC was 0.873/0.851 (95% CI 0.819–0.922/0.804–0.894). Conclusion Diabetic retinopathy (DR) lesions were detected from Optos pseudocolour UWF images using an automated algorithm. Images were classified as referral-warranted DR with a high degree of sensitivity and moderate specificity. Automated analysis of UWF images could be of value in DR screening programmes and could allow for more complete and accurate disease staging.

Telemedicine frameworks are key to screening the large, ever- growing diabetic population for preventable blindness due to diabetic retinopathy (DR). Integrating fully-automated screening systems in tele- medicine frameworks will make DR screening more ecient, cost-eective, reproducible, and accessible. In this paper, we present the integration of EyeArt, an automated DR screening system, into EyePACS, a telemedicine system for DR screening used in diverse screening settings. EyeArt in- corporates novel image processing and analysis algorithms for assessing image gradability; enhancing images based on median filtering; detecting interest regions and localizing lesions based on multi-scale morphological analysis; and DR screening and thus achieves robustness to the large im- age variability seen in a telemedicine system such as EyePACS. EyeArt is implemented as a scalable, high-throughput cloud-based system to en- able large-scale DR screening. We evaluate the safety and performance of EyeArt on a dataset with 434,023 images from 54,324 patient cases obtained from EyePACS. On this dataset, EyeArt's screening sensitivity is 90% at specificity 60.8% and the area under the receiver operating characteristic curve (AUROC) is 0.883. In a setup where trained hu- man graders review patient cases recommended for referral by EyeArt with low confidence, a workload reduction of 62% is possible. Therefore, EyeArt can be safely integrated into large real world telemedicine DR screening programs such as EyePACS helping reduce workload and in- crease eciency and thus help in reducing vision loss due to DR through early detection and treatment.

We provide a theoretical analysis of the performance of differential predictive coding using fixed-lag smoothing of the standard decoder output. This performance is compared to related results for coding using latency at the encoder, and causal encoding with delayed decoding, as well as with some prior theoretical analyses of these methods. Surprisingly, it is shown that fixed-lag smoothing of the standard decoder output with causal encoding achieves the asymptotic and finite lag performance promised by a completely reoptimized decoder.

Most video compression schemes like H.264/AVC have a high-complexity encoder with a block motion estimation (ME) engine and a low-complexity decoder. However, applications such as unmanned aerial vehicle (UAV) reconnaissance and surveillance require low-complexity video encoders. Furthermore, in such applications, the majority of the motion in the video sequences is due to the movement of the UAV and the camera mounts which is known. Motivated by this, we propose and investigate a low-complexity encoder with global motion compensation and spectral entropy based bit allocation, but without block ME. The spectral entropy based bit allocation exploits latency to look ahead at data before choosing and coding the coefficients most important for retaining signal fidelity. We show that the proposed encoder achieves better quality at lower bit rates with lower quality variation than that of the H.264 encoder with ME block size restricted to 8×8 for videos typical of UAV flyovers. Compared to the H.264 encoder with 8×8 ME blocks, the proposed encoder requires fewer computations and memory accesses.

In transform-based compression schemes, the task of choosing, quantizing, and coding the coefficients that best represent a signal is of prime importance. As a step in this direction, Yang and Gibson [1] have designed a coefficient selection scheme based on Campbell's coefficient rate and spectral entropy [2]. Building on their coefficient selection mechanism, we develop a method to allocate bits amongst the chosen coefficients that can outperform the classical method under certain conditions. We then design quantization matrices (QMs) based on the proposed bit allocation scheme. Results show that the newly designed QMs perform better than the default QMs for H.264/AVC encoding in terms of both peak signal to noise ratio (PSNR) and structural similarity (SSIM). The proposed method entails delay but is not computationally intensive.

In transform-based compression schemes, the task of choosing, quantizing, and coding the coefficients that best represent a signal is of prime importance. As a step in this direction, Yang and Gibson [1] have designed a coefficient selection scheme based on Campbell's coefficient rate and spectral entropy [2]. Building on the spectral entropy-based coefficient selection mechanism, we develop a scheme to allocate bits amongst the chosen coefficients. We show that the proposed scheme can outperform the classical method under certain conditions. We then design quantization matrices (QMs) based on the proposed bit allocation method and show that the newly designed QMs perform better than the default QMs for H.264/AVC encoding in terms of both peak signal to noise ratio (PSNR) and structural similarity (SSIM).