{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T14:23:28Z","timestamp":1787927008348,"version":"build-2784847793"},"reference-count":38,"publisher":"IOP Publishing","issue":"4","license":[{"start":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T00:00:00Z","timestamp":1787875200000},"content-version":"vor","delay-in-days":27,"URL":"https:\/\/publishingsupport.iopscience.iop.org\/iop-standard\/v1"},{"start":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T00:00:00Z","timestamp":1787875200000},"content-version":"tdm","delay-in-days":27,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Biomed. Phys. Eng. Express"],"published-print":{"date-parts":[[2026,8,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>This study investigates how architectural design influences the behavior of lightweight convolutional neural networks in multilabel classification of retinal diseases using the ocular disease intelligent recognition-5 K dataset. Rather than focusing solely on predictive performance, we analyze multiple complementary dimensions, including accuracy, stability across training runs, inter-model agreement, error complementarity, and image-level consensus. Results show that no single architecture consistently achieves both optimal performance and stability. Moreover, different architectures exhibit distinct prediction patterns and fail on partially disjoint subsets of the data, leading to significant complementarity. Agreement and consensus analyses further reveal that disagreement across models correlates with classification difficulty, suggesting its potential as a proxy for uncertainty. These findings demonstrate that architectural inductive bias plays a central role in shaping model behavior beyond aggregate metrics. Importantly, the observed diversity enables improved performance and reliability through model combination, supporting the development of ensemble-based systems for clinically relevant decision support.<\/jats:p>","DOI":"10.1088\/2057-1976\/ae987e","type":"journal-article","created":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T22:48:14Z","timestamp":1786574894000},"page":"045075","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Architectural inductive bias in from-scratch CNN training for multi-disease fundus image classification"],"prefix":"10.1088","volume":"12","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9781-3603","authenticated-orcid":true,"given":"Cris\u00f3stomo","family":"Barajas-Solano","sequence":"first","affiliation":[{"name":"Universidad de Investigaci\u00f3n y Desarrollo (UDI)"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"266","published-online":{"date-parts":[[2026,8,28]]},"reference":[{"key":"bpexae987ebib1","doi-asserted-by":"publisher","first-page":"996","DOI":"10.1109\/TMI.2020.3043495","type":"journal-article","article-title":"Modeling and enhancing low-quality retinal fundus images","volume":"40","author":"Shen","year":"2021","journal-title":"IEEE Trans. on Med. imaging"},{"key":"bpexae987ebib2","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2022.106277","type":"journal-article","article-title":"Recent trends and advances in fundus image analysis: a review","volume":"151","author":"Iqbal","year":"2022","journal-title":"Comput. Biol. Med."},{"key":"bpexae987ebib3","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2021.101971","type":"journal-article","article-title":"Applications of deep learning in fundus images: a review","volume":"69","author":"Li","year":"2021","journal-title":"Med. Image Anal."},{"key":"bpexae987ebib4","doi-asserted-by":"publisher","first-page":"354","DOI":"10.18502\/jovr.v19i3.15893","type":"journal-article","article-title":"Application of artificial intelligence in ophthalmology: an updated comprehensive review","volume":"19","author":"Hashemian","year":"2024","journal-title":"J. Ophthalmic Vis. Res."},{"key":"bpexae987ebib5","doi-asserted-by":"publisher","DOI":"10.1016\/j.inffus.2023.102059","type":"journal-article","article-title":"Fundus-deepnet: multi-label deep learning classification system for enhanced detection of multiple ocular diseases through data fusion of fundus images","volume":"102","author":"Al-Fahdawi","year":"2024","journal-title":"Inf. Fusion"},{"key":"bpexae987ebib6","doi-asserted-by":"publisher","first-page":"57796","DOI":"10.1109\/ACCESS.2022.3178372","type":"journal-article","article-title":"A comprehensive review of deep learning strategies in retinal disease diagnosis using fundus images","volume":"10","author":"Goutam","year":"2022","journal-title":"IEEE Access"},{"key":"bpexae987ebib7","article-title":"World report on vision","author":"World Health Organization","year":"2019","type":"other"},{"key":"bpexae987ebib8","first-page":"pp 10571","type":"conference-proceedings","article-title":"Attention based glaucoma detection: a large-scale database and CNN model","author":"Liu","year":"2019"},{"key":"bpexae987ebib9","doi-asserted-by":"publisher","first-page":"947","DOI":"10.3390\/mi13060947","type":"journal-article","article-title":"Multi-label fundus image classification using attention mechanisms and feature fusion","volume":"13","author":"Li","year":"2022","journal-title":"Micromachines"},{"key":"bpexae987ebib10","doi-asserted-by":"publisher","first-page":"82","DOI":"10.1186\/s12911-021-01424-x","type":"journal-article","article-title":"Multi-label classification of fundus images based on graph convolutional network","volume":"21","author":"Cheng","year":"2021","journal-title":"BMC Med. Inform. Decis. Mak."},{"key":"bpexae987ebib11","doi-asserted-by":"publisher","first-page":"195","DOI":"10.1186\/s12916-019-1426-2","type":"journal-article","article-title":"Key challenges for delivering clinical impact with artificial intelligence","volume":"17","author":"Kelly","year":"2019","journal-title":"BMC Med."},{"key":"bpexae987ebib12","doi-asserted-by":"publisher","first-page":"e537","DOI":"10.26633\/RPSP.2024.13","type":"journal-article","article-title":"Reporting guidelines for clinical trial reports for interventions involving artificial intelligence: the consort-ai extension","volume":"48","author":"Liu","year":"2024","journal-title":"Rev Panam Salud Publica"},{"key":"bpexae987ebib13","author":"Bishop","year":"2006","type":"book"},{"key":"bpexae987ebib14","doi-asserted-by":"publisher","DOI":"10.1016\/j.engappai.2022.105151","type":"journal-article","article-title":"Ensemble deep learning: a review","volume":"115","author":"Ganaie","year":"2022","journal-title":"Eng. Appl. Artif. Intell."},{"key":"bpexae987ebib15","doi-asserted-by":"publisher","DOI":"10.1016\/j.artmed.2024.102830","type":"journal-article","article-title":"Trustworthy clinical AI solutions: a unified review of uncertainty quantification in deep learning models for medical image analysis","volume":"150","author":"Lambert","year":"2024","journal-title":"Artif. Intell. Med."},{"key":"bpexae987ebib16","doi-asserted-by":"publisher","DOI":"10.1117\/1.JMI.10.5.054501","type":"journal-article","article-title":"Bayesian uncertainty estimation for detection of long-tailed and unseen conditions in medical images","volume":"10","author":"Rezaei","year":"2023","journal-title":"J. Med. Imaging"},{"key":"bpexae987ebib17","article-title":"Ocular disease intelligent recognition (odir-5k)","author":"Peking University","year":"2019","type":"other"},{"key":"bpexae987ebib18","article-title":"Ocular disease recognition (odir-5k)","author":"Andrew","year":"2020","type":"other"},{"key":"bpexae987ebib19","article-title":"Mobilenets: efficient convolutional neural networks for mobile vision applications","author":"Howard","year":"2017","type":"preprint"},{"key":"bpexae987ebib20","first-page":"pp 6105","type":"conference-proceedings","article-title":"Efficientnet: rethinking model scaling for convolutional neural networks","author":"Tan","year":"2019"},{"key":"bpexae987ebib21","first-page":"pp 113","type":"book","article-title":"4 - mobile phone-based diabetic retinopathy detection system","author":"Kashyap","year":"2020"},{"key":"bpexae987ebib22","first-page":"pp 1314","type":"conference-proceedings","article-title":"Searching for mobilenetv3","author":"Howard","year":"2019"},{"key":"bpexae987ebib23","first-page":"pp 770","type":"conference-proceedings","article-title":"Deep residual learning for image recognition","author":"Kaiming","year":"2016"},{"key":"bpexae987ebib24","first-page":"pp 4700","type":"conference-proceedings","article-title":"Densely connected convolutional networks","author":"Huang","year":"2017"},{"key":"bpexae987ebib25","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR52688.2022.01167","type":"preprint","article-title":"A convnet for the 2020s","author":"Liu","year":"2022"},{"key":"bpexae987ebib26","doi-asserted-by":"publisher","first-page":"1381","DOI":"10.32604\/iasc.2023.039315","type":"journal-article","article-title":"A consistent mistake in remote sensing images\u2019 classification literature","volume":"37","author":"Song","year":"2023","journal-title":"Intell. Autom. Soft Comput."},{"key":"bpexae987ebib27","doi-asserted-by":"publisher","first-page":"1","DOI":"10.13140\/RG.2.2.12482.66241","type":"journal-article","article-title":"A deep learning-based methodology for plant disease identification and diagnosis","volume":"18","author":"Sujawat","year":"2021","journal-title":"Webology"},{"key":"bpexae987ebib28","article-title":"PhytoCare: a hybrid approach for identifying Rice, Potato and Corn diseases","author":"Riza","year":"2023","type":"other"},{"key":"bpexae987ebib29","doi-asserted-by":"publisher","first-page":"1121","DOI":"10.1016\/j.egyr.2023.04.145","type":"journal-article","article-title":"Convnext-based anchor-free object detection model for infrared image of power equipment","volume":"9","author":"Zhang","year":"2023","journal-title":"Energy Rep."},{"key":"bpexae987ebib30","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-024-66989-9","type":"journal-article","article-title":"Enhancing agriculture through real-time grape leaf disease classification via an edge device with a lightweight cnn architecture and grad-cam","volume":"14","author":"Karim","year":"2024","journal-title":"Sci. Rep."},{"key":"bpexae987ebib31","doi-asserted-by":"publisher","first-page":"547","DOI":"10.5169\/SEALS-266450","type":"journal-article","article-title":"\u00c9tude comparative de LA distribution florale dans une portion des alpes ET du jura","volume":"37","author":"Jaccard","year":"1901","journal-title":"Bull. Soc. Vaud. Sci. Nat."},{"key":"bpexae987ebib32","doi-asserted-by":"publisher","DOI":"10.1117\/1.JMI.10.6.065503","type":"journal-article","article-title":"Generalizations of the jaccard index and s\u00f8rensen index for assessing agreement across multiple readers in object detection and instance segmentation in biomedical imaging","volume":"10","author":"Durkee","year":"2023","journal-title":"J. Med. Imaging"},{"key":"bpexae987ebib33","doi-asserted-by":"publisher","first-page":"105","DOI":"10.7320\/flmedit31si.105","type":"journal-article","article-title":"The wonder of the jaccard coefficient: from alpine floras to bipartite networks","volume":"31","author":"Podani","year":"2021","journal-title":"Flora Mediterranea"},{"key":"bpexae987ebib34","author":"Dietterich","year":"2000","type":"book"},{"key":"bpexae987ebib35","doi-asserted-by":"publisher","first-page":"379","DOI":"10.1002\/j.1538-7305.1948.tb01338.x","type":"journal-article","article-title":"A mathematical theory of communication","volume":"27","author":"Shannon","year":"1948","journal-title":"Bell Syst. Tech. J."},{"key":"bpexae987ebib36","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pdig.0000454","type":"journal-article","article-title":"BRSET: a brazilian multilabel ophthalmological dataset of retina fundus photos","volume":"3","author":"Nakayama","year":"2024","journal-title":"PLOS Digit. Health"},{"key":"bpexae987ebib37","article-title":"A brazilian multilabel ophthalmological dataset (BRSET)","author":"Luis Filipe Nakayama","year":"2023","type":"other"},{"key":"bpexae987ebib38","type":"journal-article","article-title":"Code and experimental logs for: architectural inductive bias in from-scratch CNN training for multi-disease fundus image classification","author":"Barajas-Solano","year":"2026","journal-title":"Google Drive"}],"container-title":["Biomedical Physics &amp; Engineering Express"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T13:30:39Z","timestamp":1787923839000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2057-1976\/ae987e"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,1]]},"references-count":38,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2026,8,28]]},"published-print":{"date-parts":[[2026,8,1]]}},"URL":"https:\/\/doi.org\/10.1088\/2057-1976\/ae987e","relation":{},"ISSN":["2057-1976"],"issn-type":[{"value":"2057-1976","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,1]]},"assertion":[{"value":"Architectural inductive bias in from-scratch CNN training for multi-disease fundus image classification","name":"article_title","label":"Article Title"},{"value":"Biomedical Physics & Engineering Express","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2026 IOP Publishing Ltd. All rights, including for text and data mining, AI training, and similar technologies, are reserved.","name":"copyright_information","label":"Copyright Information"},{"value":"2026-04-01","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-08-12","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-08-28","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}