{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,21]],"date-time":"2026-07-21T09:45:01Z","timestamp":1784627101194,"version":"3.55.0"},"reference-count":60,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2022,11,8]],"date-time":"2022-11-08T00:00:00Z","timestamp":1667865600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Med."],"abstract":"<jats:p>\n                    Diabetic retinopathy (DR) is a late microvascular complication of Diabetes Mellitus (DM) that could lead to permanent blindness in patients, without early detection. Although adequate management of DM\n                    <jats:italic>via<\/jats:italic>\n                    regular eye examination can preserve vision in in 98% of the DR cases, DR screening and diagnoses based on clinical lesion features devised by expert clinicians; are costly, time-consuming and not sufficiently accurate. This raises the requirements for Artificial Intelligent (AI) systems which can accurately detect DR automatically and thus preventing DR before affecting vision. Hence, such systems can help clinician experts in certain cases and aid ophthalmologists in rapid diagnoses. To address such requirements, several approaches have been proposed in the literature that use Machine Learning (ML) and Deep Learning (DL) techniques to develop such systems. However, these approaches ignore the highly valuable clinical lesion features that could contribute significantly to the accurate detection of DR. Therefore, in this study we introduce a framework called DR-detector that employs the Extreme Gradient Boosting (XGBoost) ML model trained\n                    <jats:italic>via<\/jats:italic>\n                    the combination of the features extracted by the pretrained convolutional neural networks commonly known as transfer learning (TL) models and the clinical retinal lesion features for accurate detection of DR. The retinal lesion features are extracted\n                    <jats:italic>via<\/jats:italic>\n                    image segmentation technique using the UNET DL model and captures exudates (EXs), microaneurysms (MAs), and hemorrhages (HEMs) that are relevant lesions for DR detection. The feature combination approach implemented in DR-detector has been applied to two common TL models in the literature namely VGG-16 and ResNet-50. We trained the DR-detector model using a training dataset comprising of 1,840 color fundus images collected from e-ophtha, retinal lesions and APTOS 2019 Kaggle datasets of which 920 images are healthy. To validate the DR-detector model, we test the model on external dataset that consists of 81 healthy images collected from High-Resolution Fundus (HRF) dataset and MESSIDOR-2 datasets and 81 images with DR signs collected from Indian Diabetic Retinopathy Image Dataset (IDRID) dataset annotated for DR by expert. The experimental results show that the DR-detector model achieves a testing accuracy of 100% in detecting DR after training it with the combination of ResNet-50 and lesion features and 99.38% accuracy after training it with the combination of VGG-16 and lesion features. More importantly, the results also show a higher contribution of specific lesion features toward the performance of the DR-detector model. For instance, using only the hemorrhages feature to train the model, our model achieves an accuracy of 99.38 in detecting DR, which is higher than the accuracy when training the model with the combination of all lesion features (89%) and equal to the accuracy when training the model with the combination of all lesions and VGG-16 features together. This highlights the possibility of using only the clinical features, such as lesions that are clinically interpretable, to build the next generation of robust artificial intelligence (AI) systems with great clinical interpretability for DR detection. The code of the DR-detector framework is available on GitHub at\n                    <jats:ext-link>https:\/\/github.com\/Janga-Lab\/DR-detector<\/jats:ext-link>\n                    and can be readily employed for detecting DR from retinal image datasets.\n                  <\/jats:p>","DOI":"10.3389\/fmed.2022.1050436","type":"journal-article","created":{"date-parts":[[2022,11,8]],"date-time":"2022-11-08T02:45:23Z","timestamp":1667875523000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":19,"title":["Combining transfer learning with retinal lesion features for accurate detection of diabetic retinopathy"],"prefix":"10.3389","volume":"9","author":[{"given":"Doaa","family":"Hassan","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hunter Mathias","family":"Gill","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michael","family":"Happe","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ashay D.","family":"Bhatwadekar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Amir R.","family":"Hajrasouliha","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sarath Chandra","family":"Janga","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2022,11,8]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.visres.2017.07.006","article-title":"Diabetic retinopathy, an overview","volume":"139","author":"Hartnett","year":"2017","journal-title":"Vision Res."},{"key":"B2","doi-asserted-by":"publisher","first-page":"489","DOI":"10.4239\/wjd.v6.i3.489","article-title":"Diabetic retinopathy - ocular complications of diabetes mellitus","volume":"6","author":"Nentwich","year":"2015","journal-title":"World J Diabetes."},{"key":"B3","doi-asserted-by":"publisher","first-page":"164","DOI":"10.3390\/medicina56040164","article-title":"Prevalence of diabetic retinopathy in type 1 and type 2 diabetes mellitus patients in North-East Poland","volume":"56","author":"Matuszewski","year":"2020","journal-title":"Medicina"},{"key":"B4","doi-asserted-by":"publisher","first-page":"1315","DOI":"10.2337\/dc09-2278","article-title":"Age at onset and the risk of proliferative retinopathy in type 1 diabetes","volume":"33","author":"Hietala","year":"2010","journal-title":"Diabetes Care."},{"key":"B5","doi-asserted-by":"publisher","first-page":"383","DOI":"10.1016\/bs.pmbts.2015.06.001","article-title":"Introduction to the retina","volume":"134","author":"Grossniklaus","year":"2015","journal-title":"Prog Mol Biol Transl Sci."},{"key":"B6","doi-asserted-by":"publisher","first-page":"1563","DOI":"10.1016\/j.survophthal.2022.07.008","article-title":"Role of inflammatory cells in pathophysiology and management of diabetic retinopathy","volume":"67","author":"Kovoor","year":"2022","journal-title":"Surv Ophthalmol"},{"key":"B7","doi-asserted-by":"publisher","first-page":"312","DOI":"10.4239\/wjd.v6.i2.312","article-title":"Current knowledge on diabetic retinopathy from human donor tissues","volume":"6","author":"Eisma","year":"2015","journal-title":"World J Diabetes."},{"key":"B8","doi-asserted-by":"publisher","first-page":"6669","DOI":"10.1167\/iovs.16-21031","article-title":"Diabetic retinopathy: battling the global epidemic","volume":"57","author":"Das","year":"2016","journal-title":"Invest Ophthalmol Vis Sci"},{"key":"B9","author":"Alban","year":"2016","journal-title":"Automated Detection of Diabetic Retinopathy Using Fluorescein Angiography Photographs"},{"key":"B10","doi-asserted-by":"publisher","first-page":"1122","DOI":"10.1016\/j.cell.2018.02.010","article-title":"Identifying medical diagnoses and treatable diseases by image-based deep learning","volume":"172","author":"Kermany","year":"2018","journal-title":"Cell."},{"key":"B11","unstructured":"Automated detection of diabetic retinopathy using deep learning14755\n            LamC\n            YiD\n            GuoM\n            LindseyT\n          AMIA Jt Summits Transl Sci Proc20182018"},{"key":"B12","doi-asserted-by":"publisher","first-page":"012033","DOI":"10.1088\/1742-6596\/1767\/1\/012033","article-title":"Transfer learning approach for grading of Diabetic Retinopathy","volume":"1767","author":"Aswathi","year":"2021","journal-title":"J Phys"},{"key":"B13","doi-asserted-by":"publisher","first-page":"84","DOI":"10.1145\/3065386","article-title":"ImageNet classification with deep convolutional neural networks","volume":"60","author":"Krizhevsky","year":"2017","journal-title":"Commun. ACM."},{"key":"B14","doi-asserted-by":"publisher","first-page":"41","DOI":"10.1007\/s13534-017-0047-y","article-title":"Deep-learning-based automatic computer-aided diagnosis system for diabetic retinopathy","volume":"8","author":"Mansour","year":"2018","journal-title":"Biomed Eng Lett"},{"key":"B15","article-title":"Very deep convolutional networks for large scale image recognition","author":"Simonyan","year":"2015","journal-title":"Proceedings of ICLR"},{"key":"B16","doi-asserted-by":"publisher","first-page":"1","DOI":"10.3390\/sym11010001","article-title":"Fundus image classification using VGG-19 architecture with PCA and SVD","volume":"11","author":"Mateen","year":"2019","journal-title":"Symmetry"},{"key":"B17","doi-asserted-by":"publisher","first-page":"3883","DOI":"10.3390\/s21113883","article-title":"ResNet based deep features and random forest classifier for diabetic retinopathy detection","volume":"21","author":"Yaqoob","year":"2021","journal-title":"Sensors."},{"key":"B18","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2016.90","article-title":"Deep residual learning for image recognition","volume-title":"Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR)","author":"He","year":"2016"},{"key":"B19","doi-asserted-by":"publisher","first-page":"923","DOI":"10.1007\/s11760-020-01816-y","article-title":"Deep convolution feature aggregation: an application to diabetic retinopathy severity level prediction","volume":"15","author":"Bodapati","year":"2021","journal-title":"Signal Image Video Process"},{"key":"B20","doi-asserted-by":"publisher","first-page":"914","DOI":"10.3390\/electronics9060914","article-title":"Blended multi-modal deep convnet features for diabetic retinopathy severity prediction","volume":"9","author":"Bodapati","year":"2020","journal-title":"Electronics."},{"key":"B21","doi-asserted-by":"publisher","first-page":"2369","DOI":"10.1109\/TMI.2016.2546227","article-title":"Krawiec K. Segmenting retinal blood vessels with deep neural networks","volume":"35","author":"Liskowski","year":"2016","journal-title":"IEEE Trans Med Imaging."},{"key":"B22","doi-asserted-by":"crossref","first-page":"7403","DOI":"10.1109\/ICPR48806.2021.9412088","article-title":"Learn to segment retinal lesions and beyond","volume-title":"Proceedings of 2020 25th International Conference on Pattern Recognition (ICPR)","author":"Wei","year":"2021"},{"key":"B23","doi-asserted-by":"publisher","first-page":"3","DOI":"10.4103\/0971-6203.58777","article-title":"Automated medical image segmentation techniques","volume":"35","author":"Sharma","year":"2010","journal-title":"J Med Phys."},{"key":"B24","doi-asserted-by":"publisher","first-page":"1224","DOI":"10.3390\/su13031224","article-title":"A review of deep-learning-based medical image segmentation methods","volume":"13","author":"Liu","year":"2021","journal-title":"Sustainability."},{"key":"B25","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-319-24574-4_28","article-title":"U-net: convolutional networks for biomedical image segmentation","volume-title":"Medical Image Computing and Computer-Assisted Intervention \u2013 MICCAI 2015","author":"Ronneberger","year":"2015"},{"key":"B26","doi-asserted-by":"publisher","first-page":"5327","DOI":"10.3390\/s21165327","article-title":"Neovascularization detection and localization in fundus images using deep learning","volume":"21","author":"Tang","year":"2021","journal-title":"Sensors"},{"key":"B27","doi-asserted-by":"crossref","first-page":"136","DOI":"10.1109\/ICIINFS.2018.8721315","article-title":"Diabetic retinopathy detection by retinal segmentation with region merging using CNN","volume-title":"2018 IEEE 13th International Conference on Industrial and Information Systems (ICIIS)","author":"Burewar","year":"2018"},{"key":"B28","doi-asserted-by":"publisher","first-page":"025008","DOI":"10.1117\/1.JMI.6.2.025008","article-title":"Microaneurysms segmentation with a U-Net based on recurrent residual convolutional neural network","volume":"6","author":"Kou","year":"2019","journal-title":"J Med Imaging."},{"key":"B29","doi-asserted-by":"publisher","first-page":"275","DOI":"10.3233\/XST-211073","article-title":"Automated detection of diabetic retinopathy using custom convolutional neural network","volume":"30","author":"Albahli","year":"2022","journal-title":"J Xray Sci Technol"},{"key":"B30","doi-asserted-by":"publisher","first-page":"304","DOI":"10.1007\/s10916-019-1432-0","article-title":"An intelligent segmentation and diagnosis method for diabetic retinopathy based on improved U-NET network","volume":"43","author":"Li","year":"2019","journal-title":"J Med Syst."},{"key":"B31","doi-asserted-by":"publisher","first-page":"3777","DOI":"10.3390\/app10113777","article-title":"Multi-path recurrent U-Net segmentation of retinal fundus image","volume":"10","author":"Jiang","year":"2020","journal-title":"Appl Sci"},{"key":"B32","doi-asserted-by":"publisher","first-page":"185514","DOI":"10.1109\/ACCESS.2020.3029117","article-title":"An enhanced residual U-Net for microaneurysms and exudates segmentation in fundus images","volume":"8","author":"Kou","year":"2020","journal-title":"IEEE Access"},{"key":"B33","doi-asserted-by":"publisher","first-page":"85","DOI":"10.1007\/s10916-014-0085-2","article-title":"Retinal blood vessel segmentation with neural network by using gray-level co-occurrence matrix-based features","volume":"38","author":"Rahebi","year":"2014","journal-title":"J Med Syst."},{"key":"B34","doi-asserted-by":"publisher","first-page":"6644071","DOI":"10.1155\/2021\/6644071","article-title":"FFU-Net: Feature Fusion U-Net for lesion segmentation of diabetic retinopathy","volume":"2021","author":"Xu","year":"2021","journal-title":"BioMed Res Int"},{"key":"B35","doi-asserted-by":"publisher","first-page":"167225","DOI":"10.1109\/ACCESS.2020.3023273","article-title":"U-net based method for automatic hard exudates segmentation in fundus images using inception module and residual connection","volume":"8","author":"Zong","year":"2020","journal-title":"IEEE Access"},{"key":"B36","doi-asserted-by":"publisher","first-page":"2250004","DOI":"10.1142\/S1469026822500043","article-title":"The augmentation data of retina image for blood vessel segmentation using U-Net convolutional neural network method","volume":"21","author":"Erwin","year":"2022","journal-title":"Int J Comput Intell Appl"},{"key":"B37","first-page":"2699","article-title":"Automatic screening of fundus images using a combination of convolutional neural network and hand-crafted features","volume-title":"2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), IEEE","author":"Harangi","year":"2019"},{"key":"B38","doi-asserted-by":"publisher","first-page":"103685","DOI":"10.1016\/j.bspc.2022.103685","article-title":"Enhancing CNNs through the use of hand-crafted features in automated fundus image classification","volume":"76","author":"Bogacsovics","year":"2022","journal-title":"Biomed Signal Process Control."},{"key":"B39","doi-asserted-by":"crossref","DOI":"10.1145\/2939672.2939785","article-title":"XGBoost: a scalable tree boosting system","volume-title":"Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD '16)","author":"Chen","year":"2016"},{"key":"B40","doi-asserted-by":"publisher","first-page":"248","DOI":"10.1109\/CVPR.2009.5206848","article-title":"ImageNet: A large-scale hierarchical image database","author":"Deng","year":"2009","journal-title":"Proceedings of 2009 IEEE Conference on Computer Vision and Pattern Recognition (IEEE)"},{"key":"B41","doi-asserted-by":"publisher","DOI":"10.1186\/s12886-018-0954-4","article-title":"Fundus images analysis using deep features for detection of exudates, hemorrhages and microaneurysms","author":"Khojasteh","year":"2018","journal-title":"BMC Ophthalmol"},{"key":"B42","first-page":"196","article-title":"Learn to segment retinal lesions and beyond","volume":"34","author":"Wei","year":"2013","journal-title":"Proceedings of International Conference on Pattern Recognition (ICPR)"},{"key":"B43","unstructured":"Retinal Lesions"},{"key":"B44","unstructured":"Aptos 2019- Blindness Detection"},{"key":"B45","unstructured":"High Resolution Fundus Image Database"},{"key":"B46","unstructured":"Diabetic Retinopathy \u2013 Segmentation and Grading Challenge"},{"key":"B47","unstructured":"Messidor \u2013 2"},{"key":"B48","doi-asserted-by":"publisher","first-page":"345","DOI":"10.1111\/j.1464-5491.2010.03210.x","article-title":"Retinal microaneurysm count predicts progression and regression of diabetic retinopathy. Post-hoc results from the DIRECT Programme","volume":"28","author":"Sj\u00f8lie","year":"2011","journal-title":"Diabet Med."},{"key":"B49","doi-asserted-by":"publisher","first-page":"S492","DOI":"10.4103\/2230-8210.104142","article-title":"Serum lipids and diabetic retinopathy: a cross-sectional study","volume":"16","author":"Idiculla","year":"2012","journal-title":"Indian J Endocrinol Metab."},{"key":"B50","doi-asserted-by":"publisher","first-page":"659","DOI":"10.1055\/s-0035-1556731","article-title":"Thrombosis and hemorrhage in diabetic retinopathy: a perspective from an inflammatory standpoint","volume":"41","author":"Murugesan","year":"2015","journal-title":"Semin Thromb Hemost."},{"key":"B51","doi-asserted-by":"crossref","first-page":"474","DOI":"10.1016\/B978-0-12-336156-1.50061-6","article-title":"Contrast limited adaptive histogram equalization","volume-title":"Graphics Gems IV","author":"Zuiderveld","year":"1994"},{"key":"B52","unstructured":"tf.keras.preprocessing.image.ImageDateGenerator"},{"key":"B53","unstructured":"Canny Edge Detection- OpenCV"},{"key":"B54","unstructured":"Contour Detection Using OpenCV (Python\/C++)"},{"key":"B55","unstructured":"Keras: The Python Deep Learning API"},{"key":"B56","unstructured":"Tensorflow"},{"key":"B57","doi-asserted-by":"publisher","first-page":"1145","DOI":"10.1016\/S0031-3203(96)00142-2","article-title":"Bradley. The use of the area under the roc curve in the evaluation of machine learning algorithms","volume":"30","author":"Andrew","year":"1997","journal-title":"Pattern Recogn"},{"key":"B58","unstructured":"Receiver Operating Characteristic"},{"key":"B59","unstructured":"JainA\n          Complete Guide to Parameter Tuning in XGBoost With Codes in Python2016"},{"key":"B60","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s11042-022-12642-4","article-title":"A critical review on diagnosis of diabetic retinopathy using machine learning and deep learning","volume":"23","author":"Das","year":"2022","journal-title":"Multimed Tools Appl."}],"container-title":["Frontiers in Medicine"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2022.1050436\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,11,8]],"date-time":"2022-11-08T02:45:42Z","timestamp":1667875542000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2022.1050436\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,11,8]]},"references-count":60,"alternative-id":["10.3389\/fmed.2022.1050436"],"URL":"https:\/\/doi.org\/10.3389\/fmed.2022.1050436","relation":{"has-preprint":[{"id-type":"doi","id":"10.1101\/2022.09.23.22280273","asserted-by":"object"}]},"ISSN":["2296-858X"],"issn-type":[{"value":"2296-858X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,11,8]]},"article-number":"1050436"}}