Screening for Diabetic Retinopathy Using an Automated Diagnostic System Based on Deep Learning: Diagnostic Accuracy Assessment

Rêgo, Sílvia; Dutra-Medeiros, Marco; Soares, Filipe; Monteiro‐Soares, Matilde

doi:10.1159/000512638

Cited by 19 publications

(12 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, when the fundus photos are sent to the reading centre, the image quality evaluation should precede the classification of severity of DR. Poor image quality could be due to the opacity of the refractive media, artefacts, poor contrast, defocus or small pupil 27. Previous studies assigned these ungradable pictures to referable DR,9 10 28–30 which can cause unnecessary worries to patients and confuse the graders on their judgement of referable DR or rephotography. Second, maculopathy gradability should be evaluated before grading the maculopathy.…”

Section: Discussionmentioning

confidence: 99%

Automated multidimensional deep learning platform for referable diabetic retinopathy detection: a multicentre, retrospective study

Lin

Wang

et al. 2022

BMJ Open

View full text Add to dashboard Cite

ObjectiveTo develop and validate a real-world screening, guideline-based deep learning (DL) system for referable diabetic retinopathy (DR) detection.DesignThis is a multicentre platform development study based on retrospective, cross-sectional data sets. Images were labelled by two-level certificated graders as the ground truth. According to the UK DR screening guideline, a DL model based on colour retinal images with five-dimensional classifiers, namely image quality, retinopathy, maculopathy gradability, maculopathy and photocoagulation, was developed. Referable decisions were generated by integrating the output of all classifiers and reported at the image, eye and patient level. The performance of the DL was compared with DR experts.SettingDR screening programmes from three hospitals and the Lifeline Express Diabetic Retinopathy Screening Program in China.Participants83 465 images of 39 836 eyes from 21 716 patients were annotated, of which 53 211 images were used as the development set and 30 254 images were used as the external validation set, split based on centre and period.Main outcomesAccuracy, F1 score, sensitivity, specificity, area under the receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC), Cohen’s unweighted κ and Gwet’s AC1 were calculated to evaluate the performance of the DL algorithm.ResultsIn the external validation set, the five classifiers achieved an accuracy of 0.915–0.980, F1 score of 0.682–0.966, sensitivity of 0.917–0.978, specificity of 0.907–0.981, AUROC of 0.9639–0.9944 and AUPRC of 0.7504–0.9949. Referable DR at three levels was detected with an accuracy of 0.918–0.967, F1 score of 0.822–0.918, sensitivity of 0.970–0.971, specificity of 0.905–0.967, AUROC of 0.9848–0.9931 and AUPRC of 0.9527–0.9760. With reference to the ground truth, the DL system showed comparable performance (Cohen’s κ: 0.86–0.93; Gwet’s AC1: 0.89–0.94) with three DR experts (Cohen’s κ: 0.89–0.96; Gwet’s AC1: 0.91–0.97) in detecting referable lesions.ConclusionsThe automatic DL system for detection of referable DR based on the UK guideline could achieve high accuracy in multidimensional classifications. It is suitable for large-scale, real-world DR screening.

show abstract

Section: Discussionmentioning

confidence: 99%

Automated multidimensional deep learning platform for referable diabetic retinopathy detection: a multicentre, retrospective study

Lin

Wang

et al. 2022

BMJ Open

View full text Add to dashboard Cite

show abstract

“…Progress in the application of artificial intelligence to the measurement of the FAZ area of DR [ 29 – 32 ] has allowed to mark the FAZ area, as the first step in the present study. OCTA images contain a large amount of data, which precludes manual marking due to time and personnel requirements.…”

Section: Discussionmentioning

confidence: 99%

Application of Improved U-Net Convolutional Neural Network for Automatic Quantification of the Foveal Avascular Zone in Diabetic Macular Ischemia

Meng

Lan

et al. 2022

Journal of Diabetes Research

View full text Add to dashboard Cite

Objectives.The foveal avascular zone (FAZ) is a biomarker for quantifying diabetic macular ischemia (DMI), to automate the identification and quantification of the FAZ in DMI, using an improved U-Net convolutional neural network (CNN) and to establish a CNN model based on optical coherence tomography angiography (OCTA) images for the same purpose. Methods. The FAZ boundaries on the full-thickness retina of 6 × 6 mm en face OCTA images of DMI and normal eyes were manually marked. Seventy percent of OCTA images were used as the training set, and ten percent of these images were used as the validation set to train the improved U-Net CNN with two attention modules. Finally, twenty percent of the OCTA images were used as the test set to evaluate the accuracy of this model relative to that of the baseline U-Net model. This model was then applied to the public data set sFAZ to compare its effectiveness with existing models at identifying and quantifying the FAZ area. Results. This study included 110 OCTA images. The Dice score of the FAZ area predicted by the proposed method was 0.949, the Jaccard index was 0.912, and the area correlation coefficient was 0.996. The corresponding values for the baseline U-Net were 0.940, 0.898, and 0.995, respectively, and those based on the description data set sFAZ were 0.983, 0.968, and 0.950, respectively, which were better than those previously reported based on this data set. Conclusions. The improved U-Net CNN was more accurate at automatically measuring the FAZ area on the OCTA images than the traditional CNN. The present model may measure the DMI index more accurately, thereby assisting in the diagnosis and prognosis of retinal vascular diseases such as diabetic retinopathy.

show abstract

“…The sample size (n = 578) was calculated on the basis of 95% CIs, 10% margin of error, expected sensitivity and specificity-81% and 96%, respectively [45]-and the prevalence of DR among people with diabetes of 13% [51]. Then, we inflated the sample size by an additional 25% (n = 116, total n = 462) to take account of ungradable images.…”

Section: Eligibility Criteria Study Setting Recruitment and Sample Sizementioning

confidence: 99%

“…The study was conducted with patients with known diabetic retinopathy at an ophthalmology outpatient clinic, using indirect ophthalmoscopy as the comparator. In another study, the same AI software was used to classify retinal images taken with tabletop fundus cameras [45]. Therefore, a comparison with the reference standard for screening (tabletop fundus camera) in a generalizable sample of individuals with diabetes, to assess the ability to discriminate patients with diabetic retinopathy from those without it [46] is lacking.…”

Section: Introduction 1background and Rationalementioning

confidence: 99%

Implementation and Evaluation of a Mobile Retinal Image Acquisition System for Screening Diabetic Retinopathy: Study Protocol

et al. 2022

Self Cite

View full text Add to dashboard Cite

Screening diabetic retinopathy, a major cause of blindness, is time-consuming for ophthalmologists and has some constrains in achieving full coverage and attendance. The handheld fundus camera EyeFundusScope was recently developed to expand the scale of screening, drawing on images acquired in primary care and telescreening made by ophthalmologists or a computer-aided diagnosis (CADx) system. This study aims to assess the diagnostic accuracy of the interpretation of images captured using EyeFundusScope and perform its technical evaluation, including image quality, functionality, usability, and acceptance in a real-world clinical setting. Physicians and nurses without training in ophthalmology will use EyeFundusScope to take pictures of the retinas of patients with diabetes and the images will be classified for the presence or absence of diabetic retinopathy and image quality by a panel of ophthalmologists. A subgroup of patients will also be examined with the reference standard tabletop fundus camera. Screening results provided by the CADx system on images taken with EyeFundusScope will be compared against the ophthalmologists’ analysis of images taken with the tabletop fundus camera. Diagnostic accuracy measures with 95% confidence intervals (CIs) will be calculated for positive and negative test results. Proportion of each category of image quality will be presented. Usability and acceptance results will be presented qualitatively.

show abstract

Screening for Diabetic Retinopathy Using an Automated Diagnostic System Based on Deep Learning: Diagnostic Accuracy Assessment

Cited by 19 publications

References 29 publications

Automated multidimensional deep learning platform for referable diabetic retinopathy detection: a multicentre, retrospective study

Automated multidimensional deep learning platform for referable diabetic retinopathy detection: a multicentre, retrospective study

Application of Improved U-Net Convolutional Neural Network for Automatic Quantification of the Foveal Avascular Zone in Diabetic Macular Ischemia

Implementation and Evaluation of a Mobile Retinal Image Acquisition System for Screening Diabetic Retinopathy: Study Protocol

Contact Info

Product

Resources

About