Comparison of 21 artificial intelligence algorithms in automated diabetic retinopathy screening using handheld fundus camera.

BACKGROUND: Diabetic retinopathy (DR) is a common complication of diabetes and may lead to irreversible visual loss. Efficient screening and improved treatment of both diabetes and DR have amended visual prognosis for DR. The number of patients with diabetes is increasing and telemedicine, mobile handheld devices and automated solutions may alleviate the burden for healthcare. We compared the performance of 21 artificial intelligence (AI) algorithms for referable DR screening in datasets taken by handheld Optomed Aurora fundus camera in a real-world setting. PATIENTS AND METHODS: Prospective study of 156 patients (312 eyes) attending DR screening and follow-up. Both papilla- and macula-centred 50° fundus images were taken from each eye. DR was graded by experienced ophthalmologists and 21 AI algorithms. RESULTS: Most eyes, 183 out of 312 (58.7%), had no DR and mild NPDR was noted in 21 (6.7%) of the eyes. Moderate NPDR was detected in 66 (21.2%) of the eyes, severe NPDR in 1 (0.3%), and PDR in 41 (13.1%) composing a group of 34.6% of eyes with referable DR. The AI algorithms achieved a mean agreement of 79.4% for referable DR, but the results varied from 49.4% to 92.3%. The mean sensitivity for referable DR was 77.5% (95% CI 69.1-85.8) and specificity 80.6% (95% CI 72.1-89.2). The rate for images ungradable by AI varied from 0% to 28.2% (mean 1.9%). Nineteen out of 21 (90.5%) AI algorithms resulted in grading for DR at least in 98% of the images. CONCLUSIONS: Fundus images captured with Optomed Aurora were suitable for DR screening. The performance of the AI algorithms varied considerably emphasizing the need for external validation of screening algorithms in real-world settings before their clinical application.

What is already known on this topic? Diabetic retinopathy (DR) is a common complication of diabetes. Efficient screening and timely treatment are important to avoid the development of sight-threatening DR. The increasing number of patients with diabetes and DR poses a challenge for healthcare.What this study adds? Telemedicine, mobile handheld devices and artificial intelligence (AI)-based automated algorithms are likely to alleviate the burden by improving efficacy of DR screening programs. Reliable algorithms of high quality exist despite the variability between the solutions.How this study might affect research, practice or policy? AI algorithms improve the efficacy of screening and might be implemented to clinical use after thorough validation in a real-life setting.

Rotational Stability, Tilt and Decentration of a New IOL with a 7.0 mm Optic.

Purpose: To evaluate rotational stability, tilt and decentration of a new monofocal IOL with a 7.0 mm optic and frame haptics.Methods: Prospective post-market clinical follow-up study at the Kepler University Hospital Linz, Austria. An Aspira-aXA (HumanOptics, Germany) was implanted in 74 eyes of 42 cataract patients. The lens was manufactured with toric markings. IOL rotational stability was evaluated by comparing its position at the end of surgery (EoS) versus 1 day, 1 week, 1 month and 4 months postoperatively. IOL tilt and decentration were measured using a Scheimpflug camera at 1 week, 1 month and 4 months.Results: Median absolute IOL rotation was 1.42 degrees (n = 52; mean = 2.18 ± 2.23°) within 1 day after surgery and was significantly higher compared to all later intervals (median <1.0 degree; P = .001). At the 4 months follow-up, IOL rotation was within 5.0 degrees in 85% of the eyes (n = 40) and within 10.0 degrees in 98% (n = 46) of the eyes. The only eye with an IOL rotation of ≥ 10.ty0 degrees (EoS vs. 1 day) had an AL of 26.45 mm. At the last follow-up, the IOL vertical and horizontal tilt referenced to the pupillary axis was in average less than 1.5 degrees in both eyes (n = 54; maximum 5.85°). Decentration in both meridians was on average less than 0.10 mm in both eyes (maximum 0.30 mm).Conclusion: The one-piece Aspira-aXA IOL showed good and stable positioning within the capsular bag over a 4 months period.

A Novel Virtual Wet Lab-Using a Smartphone Camera Adapter and a Video Conference Platform to Provide Real-Time Surgical Instruction.

PURPOSE: Proctored surgical instruction has traditionally been taught through in-person interactions in either the operating room or an improvised wet lab. Because of the COVID-19 pandemic, live in-person instruction was not feasible owing to social distancing protocols, so a virtual wet lab (VWL) was proposed and implemented. The purpose of this article is to describe our experience with a VWL as a Descemet membrane endothelial keratoplasty (DMEK) skills-transfer course. This is the first time that a VWL environment has been described for the instruction of ophthalmic surgery. METHODS: Thirteen participant surgeons took part in VWLs designed for DMEK skills transfer in September and October 2020. A smartphone camera adapter and a video conference software platform were the unique media for the VWL. After a didactic session, participants were divided into breakout rooms where their surgical scope view was broadcast live, allowing instructors to virtually proctor their participants in real time. Participants were surveyed to assess their satisfaction with the course. RESULTS: All (100%) participants successfully injected and unfolded their DMEK grafts. Ten of the 13 participants completed the survey. Respondents rated the experience highly favorably. CONCLUSIONS: With the use of readily available technology, VWLs can be successfully implemented in lieu of in-person skills-transfer courses. Further development catering to the needs of the participant might allow VWLs to serve as a viable option of surgical education, currently limited by geographical and social distancing boundaries.

Validation of a Three-Dimensional Head and Neck Spheroid Model to Evaluate Cameras for NIR Fluorescence-Guided Cancer Surgery.

Near-infrared (NIR) fluorescence-guided surgery is an innovative technique for the real-time visualization of resection margins. The aim of this study was to develop a head and neck multicellular tumor spheroid model and to explore the possibilities offered by it for the evaluation of cameras for NIR fluorescence-guided surgery protocols. FaDu spheroids were incubated with indocyanine green (ICG) and then included in a tissue-like phantom. To assess the capability of Fluobeam® NIR camera to detect ICG in tissues, FaDu spheroids exposed to ICG were embedded in 2, 5 or 8 mm of tissue-like phantom. The fluorescence signal was significantly higher between 2, 5 and 8 mm of depth for spheroids treated with more than 5 µg/mL ICG (p < 0.05). The fluorescence intensity positively correlated with the size of spheroids (p < 0.01), while the correlation with depth in the tissue-like phantom was strongly negative (p < 0.001). This multicellular spheroid model embedded in a tissue-like phantom seems to be a simple and reproducible in vitro tumor model, allowing a comparison of NIR cameras. The ideal configuration seems to be 450 µm FaDu spheroids incubated for 24 hours with 0.05 mg/ml of ICG, ensuring the best stability, toxicity, incorporation and signal intensity.

Monitoring social distancing under various low light conditions with deep learning and a single motionless time of flight camera.

The purpose of this work is to provide an effective social distance monitoring solution in low light environments in a pandemic situation. The raging coronavirus disease 2019 (COVID-19) caused by the SARS-CoV-2 virus has brought a global crisis with its deadly spread all over the world. In the absence of an effective treatment and vaccine the efforts to control this pandemic strictly rely on personal preventive actions, e.g., handwashing, face mask usage, environmental cleaning, and most importantly on social distancing which is the only expedient approach to cope with this situation. Low light environments can become a problem in the spread of disease because of people's night gatherings. Especially, in summers when the global temperature is at its peak, the situation can become more critical. Mostly, in cities where people have congested homes and no proper air cross-system is available. So, they find ways to get out of their homes with their families during the night to take fresh air. In such a situation, it is necessary to take effective measures to monitor the safety distance criteria to avoid more positive cases and to control the death toll. In this paper, a deep learning-based solution is proposed for the above-stated problem. The proposed framework utilizes the you only look once v4 (YOLO v4) model for real-time object detection and the social distance measuring approach is introduced with a single motionless time of flight (ToF) camera. The risk factor is indicated based on the calculated distance and safety distance violations are highlighted. Experimental results show that the proposed model exhibits good performance with 97.84% mean average precision (mAP) score and the observed mean absolute error (MAE) between actual and measured social distance values is 1.01 cm.

Habitat suitability maps for juvenile tri-spine horseshoe crabs in Japanese intertidal zones: A model approach using unmanned aerial vehicles and the Structure from Motion technique.

The tri-spine horseshoe crab, Tachypleus tridentatus, is a threatened species that inhabits coastal areas from South to East Asia. A Conservation management system is urgently required for managing its nursery habitats, i.e., intertidal flats, especially in Japan. Habitat suitability maps are useful in drafting conservation plans; however, they have rarely been prepared for juvenile T. tridentatus. In this study, we examined the possibility of constructing robust habitat suitability models (HSMs) for juveniles based on topographical data acquired using unmanned aerial vehicles and the Structure from Motion (UAV-SfM) technique. The distribution data of the juveniles in the Tsuyazaki and Imazu intertidal flats from 2017 to 2019 were determined. The data were divided into a training dataset for HSM construction and three test datasets for model evaluation. High accuracy digital surface models were built for each region using the UAV-SfM technique. Normalized elevation was assessed by converting the topographical models that consider the tidal range in each region, and the slope was calculated based on these models. Using the training data, HSMs of the juveniles were constructed with normalized elevation and slope as the predictor variables. The HSMs were evaluated using the test data. The results showed that HSMs exhibited acceptable discrimination performance for each region. Habitat suitability maps were built for the juveniles in each region, and the suitable areas were estimated to be approximately 6.1 ha of the total 19.5 ha in Tuyazaki, and 3.7 ha of the total 7.9 ha area in Imazu. In conclusion, our findings support the usefulness of the UAV-SfM technique in constructing HSMs for juvenile T. tridentatus. The monitoring of suitable habitat areas for the juveniles using the UAV-SfM technique is expected to reduce survey costs, as it can be conducted with fewer investigators over vast intertidal zones within a short period of time.

Visualization of Low-Level Gamma Radiation Sources Using a Low-Cost, High-Sensitivity, Omnidirectional Compton Camera.

We present experimental protocols for visualizing various low-level gamma radiation sources in the ambient environment. Experiments were conducted by using a low-cost, high-sensitivity, omnidirectional, gamma-ray imaging Compton camera. In the laboratory, the position of a sub-MeV gamma radiation source such as 137Cs can easily be monitored via omnidirectional gamma-ray imaging obtained by the Compton camera. In contrast, a stationary, wall-mounted dose rate monitor cannot always successfully monitor such a source. Furthermore, we successfully demonstrated the possibility of visualizing the radioactivity movement in the environment, for example, the movement of a patient injected with 18F-fluorodeoxyglucose (18F-FDG) in a nuclear medicine facility. In the Fukushima field, we easily obtained omnidirectional gamma-ray images concerned with the distribution on the ground of low-level radioactive contamination by radioactive cesium released by the Fukushima Daiichi nuclear power plant accident in 2011. We demonstrate clear advantages of using our procedure with this camera to visualize gamma-ray sources. Our protocols can further be used to discover low-level gamma radiation sources, in place of stationary dose rate monitors and/or portable survey meters used conventionally.

Comparison of Widefield Imaging Between Confocal Laser Scanning Ophthalmoscopy and Broad Line Fundus Imaging in Routine Clinical Practice.

BACKGROUND AND OBJECTIVE: The purpose of this study was to evaluate the difference between widefield confocal scanning laser imaging (WSLO) and widefield broad line fundus (WBLF) imaging in their ability to view the peripheral retina in routine clinical practice. PATIENTS AND METHODS: A retrospective chart review identified patients within routine clinical practice who were imaged with a WSLO image and a single and montaged WBLF image. The primary outcome was the number of ultra-widefield quadrants captured utilizing the UWF consensus definitions. Secondary outcomes included the area within each of quadrant and the differences in clinical grading between modalities. RESULTS: More vortex ampullae were identified with the WSLO than either single image or montage WBLF image. The WSLO captured 116 of the possible 260 vortex ampullae (45%) in comparison to the WBLF single image (8 of 260; 3%) and WBLF montage (96 of 260; 37%). Only five eyes from WSLO and no images from the WBLF single image met the ultra-widefield consensus definition in routine clinical practice. The average area per individual quadrant acquired by WSLO image was greater than the single or montage WBLF image (781.67 mm2, 433.82 mm2, and 686.03 mm2, respectively; P < .001). Clinical grading of images found a substantial inter-rater agreement with both technologies (86% on WSLO; 88% on WBLF). CONCLUSIONS: Both systems had a low rate of meeting UWF consensus definitions in routine clinical practice. A single WSLO image acquired a greater area than WBLF image in both single-image and montage formats. [Ophthalmic Surg Lasers Imaging Retina. 2020;51:89-94.].

Interobserver and intraobserver variability affecting the assessment of loss of autofluorescence of oral mucosal lesions.

OBJECTIVES: The assessment of loss of tissue autofluorescence (LAF) has been proposed as an adjunct to comprehensive oral examination to enhance the detection of mucosal lesions harbouring dysplasia or carcinoma. The assessment of LAF is not based on completely objectified parameters therefore intraobserver and interobserver variability cannot be neglected alongside the issue of correct interpretation of LAF. The present study evaluated intraobserver and interobserver variability in the clinical assessment of LAF as performed by oral medicine practitioners (OMPs) or general dental practitioners (GDPs). MATERIALS AND METHODS: Couples of clinical pictures, acquired under white incandescent dental operatory light and during the assessment of LAF performed by VELscope were retrieved. Four OMPs and eight GDPs were asked to assess the pictures and to score the LAF. Kappa statistics allowed the assessment of intra- and inter-observer related variability. RESULTS: Pictures of 109 lesions representative of all oral mucosal sites and clinical appearances were selected. OMPs had a better intraobserver agreement than GDPs (substantial versus moderate). The moderate (k = 0.506) interobserver agreement observed among both OMPs and GDPs in a 2-score model (positive versus negative), lowered down to poor values only among GDPs when a 3-score or 4-score model (including uncertain judgements) was applied. CONCLUSIONS: A good agreement (k>0.8) was never observed and the present results are similar to previously reported data about conventional oral examination. Irrespective of the diagnostic accuracy, the assessment of AF seems not to be able to improve observer-related variability in the clinical assessment of oral mucosal lesions.

Assessment of Accuracy of an Artificial Intelligence Algorithm to Detect Melanoma in Images of Skin Lesions.

Importance: A high proportion of suspicious pigmented skin lesions referred for investigation are benign. Techniques to improve the accuracy of melanoma diagnoses throughout the patient pathway are needed to reduce the pressure on secondary care and pathology services. Objective: To determine the accuracy of an artificial intelligence algorithm in identifying melanoma in dermoscopic images of lesions taken with smartphone and digital single-lens reflex (DSLR) cameras. Design, Setting, and Participants: This prospective, multicenter, single-arm, masked diagnostic trial took place in dermatology and plastic surgery clinics in 7 UK hospitals. Dermoscopic images of suspicious and control skin lesions from 514 patients with at least 1 suspicious pigmented skin lesion scheduled for biopsy were captured on 3 different cameras. Data were collected from January 2017 to July 2018. Clinicians and the Deep Ensemble for Recognition of Malignancy, a deterministic artificial intelligence algorithm trained to identify melanoma in dermoscopic images of pigmented skin lesions using deep learning techniques, assessed the likelihood of melanoma. Initial data analysis was conducted in September 2018; further analysis was conducted from February 2019 to August 2019. Interventions: Clinician and algorithmic assessment of melanoma. Main Outcomes and Measures: Area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity of the algorithmic and specialist assessment, determined using histopathology diagnosis as the criterion standard. Results: The study population of 514 patients included 279 women (55.7%) and 484 white patients (96.8%), with a mean (SD) age of 52.1 (18.6) years. A total of 1550 images of skin lesions were included in the analysis (551 [35.6%] biopsied lesions; 999 [64.4%] control lesions); 286 images (18.6%) were used to train the algorithm, and a further 849 (54.8%) images were missing or unsuitable for analysis. Of the biopsied lesions that were assessed by the algorithm and specialists, 125 (22.7%) were diagnosed as melanoma. Of these, 77 (16.7%) were used for the primary analysis. The algorithm achieved an AUROC of 90.1% (95% CI, 86.3%-94.0%) for biopsied lesions and 95.8% (95% CI, 94.1%-97.6%) for all lesions using iPhone 6s images; an AUROC of 85.8% (95% CI, 81.0%-90.7%) for biopsied lesions and 93.8% (95% CI, 91.4%-96.2%) for all lesions using Galaxy S6 images; and an AUROC of 86.9% (95% CI, 80.8%-93.0%) for biopsied lesions and 91.8% (95% CI, 87.5%-96.1%) for all lesions using DSLR camera images. At 100% sensitivity, the algorithm achieved a specificity of 64.8% with iPhone 6s images. Specialists achieved an AUROC of 77.8% (95% CI, 72.5%-81.9%) and a specificity of 69.9%. Conclusions and Relevance: In this study, the algorithm demonstrated an ability to identify melanoma from dermoscopic images of selected lesions with an accuracy similar to that of specialists.

Uso de retinógrafo portátil como ferramenta no rastreamento de retinopatia diabética / Use of a handheld fundus camera as a screening tool for diabetic retinopathy

Resumo Objetivo: Determinar a concordância do retinógrafo portátil Pictor Plus, na ausência de midríase, com a retinografia digital e avaliar a sua acurácia como método de rastreio da retinopatia diabética (RD). Métodos: Estudo transversal, mascarado para o observador, avaliando 100 olhos de 51 pacientes diabéticos. Foram realizadas retinografias com o Pictor Plus com e sem midríase, seguidos de retinografia convencional com o retinógrafo IMAGEnet por técnico treinado. As imagens obtidas foram analisadas por oftalmologista especialista em retina e classificadas normais, presença de edema macular diabético, retinopatia não proliferativa (inicial, moderada e grave) e retinopatia proliferativa, além de análise inviável. A concordância entre os resultados foi avaliada via coeficiente Kappa. As imagens foram agrupadas em normais e alteradas e estas em RD de alto risco e RD de baixo risco. Avaliou-se ainda a sensibilidade, especificidade, valores preditivos positivos e negativos, em relação à retinografia convencional. Resultados: Oitenta e nove por cento das imagens foram consideradas viáveis para classificação. Pouco mais de 31% dos olhos avaliados foram considerados como normais pelas três tecnologias. O exame com Pictor na ausência de midríase apresentou altos índices de sensibilidade e especificidade para a classificação normal (92,9% e 93,4%) e RD de alto risco (82,9% e 97,9%) e bons resultados para RD de baixo risco (75,0% e 87,0%). Conclusão: O Pictor Plus apresentou altos níveis de concordância com a retinografia digital ao discriminar exames normais de alterados e portadores de RD de baixo e alto risco. Foram verificadas ainda altas sensibilidade e especificidade desta ferramenta, se comparados à retinografia padrão.

Abstract Purpose: Determine the effectiveness of the use a portable handheld fundus camera as a screening method for diabetic retinopathy, compared to regular digital retinography. Methods: This is a cross-sectional, observer-masked study, evaluating 100 eyes of 51 diabetic patients undergoing a routine dilated retinography for diabetic retinopathy. Fundus images pre and post-dilation we captured by Pictor Plus (Volk Optical Inc., Mentor, USA), followed by routine digital retinography by IMAGEnet (Topcon Medical Systems, New Jersey, USA). All exams were performed by a trained technician on the same occasion. The images were analyzed and graded by a masked retina specialist and classified as normal, presence of diabetic macular edema, nonproliferative diabetic retinopathy (initial, moderate and severe) and proliferative diabetic retinopathy. The ungradable images were recorded and excluded from analyses. The agreement between results obtained by the three methods was evaluated via Kappa coefficient. Sensitivity, specificity, positive and negative predictive values in relation to IMAGEnet images were also determined. Results: Images were gradable in 89% of pre-dilation photos. Pictor Plus pre-dilation images had high sensitivity and specificity in identifying normal eyes (92.9% and 93.4%) and in vision-threatening DR (82.9% and 97.9%) both when compared to IMAGEnet results. Conclusion: Pictor can capture retinal images of sufficient quality to screen for DR with and without dilation. Single retinal images obtained using Pictor can identify eyes with vision-threatening DR with high sensitivity and specificity compared to routine IMAGEnet images.