{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,29]],"date-time":"2025-10-29T17:36:08Z","timestamp":1761759368657,"version":"build-2065373602"},"reference-count":39,"publisher":"Public Library of Science (PLoS)","issue":"10","license":[{"start":{"date-parts":[[2025,10,29]],"date-time":"2025-10-29T00:00:00Z","timestamp":1761696000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100001607","name":"Eye and Ear Foundation of Pittsburgh","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100001607","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100001818","name":"Research to Prevent Blindness","doi-asserted-by":"publisher","award":["Medical Student Eye Research Fellowship"],"award-info":[{"award-number":["Medical Student Eye Research Fellowship"]}],"id":[{"id":"10.13039\/100001818","id-type":"DOI","asserted-by":"publisher"}]},{"name":"NIH","award":["P30 EY08098"],"award-info":[{"award-number":["P30 EY08098"]}]},{"name":"The Hillman Challenge Grant - Exploratory Research Project"}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:sec id=\"sec001\">\n                    <jats:title>Objectives<\/jats:title>\n                    <jats:p>Retinal diseases, major causes of vision impairment and blindness, are assessed using optical coherence tomography (OCT) scans. Automated report generation for retinal OCT scans, powered by deep learning, can help standardize interpretations and track retinal disease in clinical settings. A key challenge is accurately segmenting retinal disease signatures. This study explores using the diffusion model to segment subretinal fluid (SRF), intraretinal fluid (IRF), and pigment epithelial detachment (PED) in typical clinical settings, comparing their performance to other leading segmentation models.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec002\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>\n                      We labeled OCT scans and extracted those with specific pathologic retinal features: 269 scans with SRF, 224 scans with IRF, and 114 scans with PED. Three trained reviewers manually segmented these features for downstream analysis. Using manually segmented scans as the ground truth, we trained the diffusion model, Nested U-Net, nnU-Net, TransUNet, and SwinUNet to predict these segmentations. All models were evaluated using 5-fold cross-validation, with performance measured by Dice coefficient, sensitivity, specificity, Pearson correlation coefficient, and R\n                      <jats:sup>2<\/jats:sup>\n                      .\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec003\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      All models show high similarly with ground truth segmentations in predicting SRF, IRF, and PED, as shown by the Dice coefficient (Diffusion model: 0.81\u2009\u00b1\u20090.12, 0.66\u2009\u00b1\u20090.09, 0.75\n                      <jats:bold>\u2009<\/jats:bold>\n                      \u00b1\u20090.11). The diffusion model has relatively higher sensitivity compared to most other models, while all models display very high specificity. The Pearson correlation coefficient and R\n                      <jats:sup>2<\/jats:sup>\n                      values show strongly associated pixel quantification of segmented areas for models, with the nnU-Net model performing the strongest overall.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec004\">\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>This study demonstrates that while diffusion models can comparably segment retinal pathologies using a limited number of manually annotated scans, the nnU-Net model remains the most effective overall for automated OCT analysis.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1371\/journal.pone.0335615","type":"journal-article","created":{"date-parts":[[2025,10,29]],"date-time":"2025-10-29T17:33:25Z","timestamp":1761759205000},"page":"e0335615","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Benchmarking diffusion models against state-of-the-art architectures for OCT fluid biomarker segmentation"],"prefix":"10.1371","volume":"20","author":[{"given":"Katherine","family":"Du","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Utkarsh","family":"Doshi","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Benjamin","family":"DiCenzo","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jessica","family":"Jiang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ethan","family":"Wu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Adarsh","family":"Gadari","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Sharat Chandra","family":"Vupparaboina","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Elham","family":"Sadeghi","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Sandeep Chandra","family":"Bollepalli","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jos\u00e9-Alain","family":"Sahel","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jay","family":"Chhablani","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0024-8404","authenticated-orcid":true,"given":"Kiran Kumar","family":"Vupparaboina","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"340","published-online":{"date-parts":[[2025,10,29]]},"reference":[{"issue":"11","key":"pone.0335615.ref001","doi-asserted-by":"crossref","first-page":"3283","DOI":"10.1007\/s00417-023-06052-x","article-title":"Artificial intelligence in retinal disease: clinical application, challenges, and future directions","volume":"261","author":"M Daich Varela","year":"2023","journal-title":"Graefes Arch Clin Exp Ophthalmol"},{"issue":"12","key":"pone.0335615.ref002","doi-asserted-by":"crossref","first-page":"2489","DOI":"10.1016\/j.ophtha.2014.07.054","article-title":"Retinal imaging in the twenty-first century: state of the art and future directions","volume":"121","author":"PA Keane","year":"2014","journal-title":"Ophthalmology"},{"issue":"2","key":"pone.0335615.ref003","first-page":"247","article-title":"Classification of retinal diseases based on OCT images","volume":"23","author":"N Eladawi","year":"2018","journal-title":"Front Biosci"},{"issue":"13","key":"pone.0335615.ref004","doi-asserted-by":"crossref","first-page":"5200","DOI":"10.1167\/iovs.16-19964","article-title":"Improved automated detection of diabetic retinopathy on a publicly available dataset through integration of deep learning","volume":"57","author":"MD Abr\u00e0moff","year":"2016","journal-title":"Invest Ophthalmol Vis Sci"},{"issue":"22","key":"pone.0335615.ref005","doi-asserted-by":"crossref","first-page":"2402","DOI":"10.1001\/jama.2016.17216","article-title":"Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs","volume":"316","author":"V Gulshan","year":"2016","journal-title":"JAMA"},{"issue":"4","key":"pone.0335615.ref006","doi-asserted-by":"crossref","first-page":"322","DOI":"10.1016\/j.oret.2016.12.009","article-title":"Deep learning is effective for the classification of OCT images of normal versus age-related macular degeneration","volume":"1","author":"CS Lee","year":"2017","journal-title":"Ophthalmol Retina"},{"issue":"9","key":"pone.0335615.ref007","doi-asserted-by":"crossref","first-page":"1342","DOI":"10.1038\/s41591-018-0107-6","article-title":"Clinically applicable deep learning for diagnosis and referral in retinal disease","volume":"24","author":"J De Fauw","year":"2018","journal-title":"Nat Med"},{"issue":"1","key":"pone.0335615.ref008","doi-asserted-by":"crossref","first-page":"19545","DOI":"10.1038\/s41598-023-46626-7","article-title":"Automated deep learning-based AMD detection and staging in real-world OCT datasets (PINNACLE study report 5)","volume":"13","author":"O Leingang","year":"2023","journal-title":"Scientific Reports"},{"issue":"5","key":"pone.0335615.ref009","doi-asserted-by":"crossref","first-page":"693","DOI":"10.1016\/j.ophtha.2020.09.025","article-title":"Quantitative analysis of OCT for neovascular age-related macular degeneration using deep learning","volume":"128","author":"G Moraes","year":"2021","journal-title":"Ophthalmology"},{"issue":"1","key":"pone.0335615.ref010","doi-asserted-by":"crossref","first-page":"13605","DOI":"10.1038\/s41598-019-49740-7","article-title":"Expert-level automated biomarker identification in optical coherence tomography scans","volume":"9","author":"T Kurmann","year":"2019","journal-title":"Sci Rep"},{"issue":"4","key":"pone.0335615.ref011","doi-asserted-by":"crossref","first-page":"407","DOI":"10.3390\/bioengineering10040407","article-title":"On machine learning in clinical interpretation of retinal diseases using OCT images","volume":"10","author":"PK Karn","year":"2023","journal-title":"Bioengineering (Basel)"},{"key":"pone.0335615.ref012","doi-asserted-by":"crossref","first-page":"91","DOI":"10.1016\/j.cmpb.2019.06.019","article-title":"Automatic segmentation of hyperreflective foci in OCT images","volume":"178","author":"L Varga","year":"2019","journal-title":"Comp Method Prog Biomed"},{"issue":"10","key":"pone.0335615.ref013","doi-asserted-by":"crossref","first-page":"1436","DOI":"10.1136\/bjo-2023-324647","article-title":"Validation of a deep learning model for automatic detection and quantification of five OCT critical retinal features associated with neovascular age-related macular degeneration","volume":"108","author":"F Ricardi","year":"2024","journal-title":"Br J Ophthalmol"},{"issue":"4","key":"pone.0335615.ref014","doi-asserted-by":"crossref","first-page":"17","DOI":"10.1167\/tvst.10.4.17","article-title":"Automated quantification of pathological fluids in neovascular age-related macular degeneration, and its repeatability using deep learning","volume":"10","author":"I Mantel","year":"2021","journal-title":"Transl Vis Sci Technol"},{"issue":"4","key":"pone.0335615.ref015","doi-asserted-by":"crossref","first-page":"549","DOI":"10.1016\/j.ophtha.2017.10.031","article-title":"Fully automated detection and quantification of macular fluid in OCT using deep learning","volume":"125","author":"T Schlegl","year":"2018","journal-title":"Ophthalmology"},{"issue":"18","key":"pone.0335615.ref016","doi-asserted-by":"crossref","first-page":"3793","DOI":"10.1038\/s41433-023-02615-8","article-title":"Performance of retinal fluid monitoring in OCT imaging by automated deep learning versus human expert grading in neovascular AMD","volume":"37","author":"M Pawloff","year":"2023","journal-title":"Eye (Lond)"},{"issue":"1","key":"pone.0335615.ref017","doi-asserted-by":"crossref","first-page":"517","DOI":"10.1038\/s41598-023-27616-1","article-title":"A vision transformer architecture for the automated segmentation of retinal lesions in spectral domain optical coherence tomography images","volume":"13","author":"D Philippi","year":"2023","journal-title":"Sci Rep"},{"issue":"9","key":"pone.0335615.ref018","doi-asserted-by":"crossref","first-page":"1350","DOI":"10.1136\/bjophthalmol-2022-321348","article-title":"Automatic segmentation of multitype retinal fluid from optical coherence tomography images using semisupervised deep learning network","volume":"107","author":"F Li","year":"2023","journal-title":"Br J Ophthalmol"},{"issue":"4","key":"pone.0335615.ref019","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3626235","article-title":"Diffusion models: a comprehensive survey of methods and applications","volume":"56","author":"L Yang","year":"2023","journal-title":"ACM Comput Surv"},{"key":"pone.0335615.ref020","article-title":"DiffusionDCI: a novel diffusion-based unified framework for dynamic full-field OCT image generation and segmentation","author":"B Yang","year":"2024","journal-title":"IEEE Access"},{"key":"pone.0335615.ref021","first-page":"1","article-title":"Retinal OCT synthesis with denoising diffusion probabilistic models for layer segmentation.","author":"Y Wu","year":"2024","journal-title":"2024 IEEE International symposium on biomedical imaging (ISBI)"},{"key":"pone.0335615.ref022","first-page":"25","article-title":"Unsupervised denoising of retinal OCT with diffusion probabilistic model.","volume-title":"Medical imaging 2022: image processing","author":"D Hu","year":"2022"},{"key":"pone.0335615.ref023","first-page":"660","article-title":"Content-preserving diffusion model for unsupervised as-oct image despeckling.","volume-title":"International conference on medical image computing and computer-assisted intervention","author":"S Li","year":"2023"},{"key":"pone.0335615.ref024","first-page":"284","article-title":"A novel diffusion-model-based OCT image inpainting algorithm for wide saturation artifacts.","volume-title":"Chinese conference on pattern recognition and computer vision (PRCV)","author":"B Ji","year":"2023"},{"issue":"12","key":"pone.0335615.ref025","article-title":"Inter-rater reliability in labeling quality and pathological features of retinal OCT scans: a customized annotation software approach","volume":"19","author":"K Du","year":"2024","journal-title":"PLoS One"},{"key":"pone.0335615.ref026","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1007\/978-3-030-00889-5_1","article-title":"Unet: A nested u-net architecture for medical image segmentation.","volume-title":"Deep learning in medical image analysis and multimodal learning for clinical decision support: 4th international workshop, DLMIA 2018, and 8th international workshop, ML-CDS 2018, held in conjunction with MICCAI 2018, Granada, Spain, September 20, 2018, proceedings","author":"Z Zhou","year":"2018"},{"issue":"2","key":"pone.0335615.ref027","doi-asserted-by":"crossref","first-page":"203","DOI":"10.1038\/s41592-020-01008-z","article-title":"nnU-Net: a self-configuring method for deep learning-based biomedical image segmentation","volume":"18","author":"F Isensee","year":"2021","journal-title":"Nat Methods"},{"key":"pone.0335615.ref028","unstructured":"Chen J, Lu Y, Yu Q, Luo X, Adeli E, Wang Y, et al. Transunet: transformers make strong encoders for medical image segmentation. 2021. https:\/\/arxiv.org\/abs\/2102.04306"},{"key":"pone.0335615.ref029","first-page":"205","article-title":"Swin-unet: Unet-like pure transformer for medical image segmentation.","volume-title":"European conference on computer vision","author":"H Cao","year":"2022"},{"key":"pone.0335615.ref030","first-page":"6840","article-title":"Denoising diffusion probabilistic models","volume":"33","author":"J Ho","year":"2020","journal-title":"Adv Neural Inform Process Syst"},{"key":"pone.0335615.ref031","first-page":"2256","article-title":"Deep unsupervised learning using nonequilibrium thermodynamics.","volume-title":"International conference on machine learning","author":"J Sohl-Dickstein","year":"2015"},{"key":"pone.0335615.ref032","first-page":"1623","article-title":"Medsegdiff: medical image segmentation with diffusion probabilistic model.","volume-title":"Medical imaging with deep learning","author":"J Wu","year":"2024"},{"issue":"3","key":"pone.0335615.ref033","doi-asserted-by":"crossref","first-page":"297","DOI":"10.2307\/1932409","article-title":"Measures of the amount of ecologic association between species","volume":"26","author":"LR Dice","year":"1945","journal-title":"Ecology"},{"issue":"1","key":"pone.0335615.ref034","doi-asserted-by":"crossref","first-page":"45","DOI":"10.4103\/0301-4738.37595","article-title":"Understanding and using sensitivity, specificity and predictive values","volume":"56","author":"R Parikh","year":"2008","journal-title":"Indian J Ophthalmol"},{"issue":"5","key":"pone.0335615.ref035","doi-asserted-by":"crossref","first-page":"1763","DOI":"10.1213\/ANE.0000000000002864","article-title":"Correlation coefficients: appropriate use and interpretation","volume":"126","author":"P Schober","year":"2018","journal-title":"Anesth Analg"},{"issue":"1","key":"pone.0335615.ref036","doi-asserted-by":"crossref","DOI":"10.1049\/ipr2.70019","article-title":"A comprehensive review of U\u2010Net and its variants: advances and applications in medical image segmentation","volume":"19","author":"W Jiangtao","year":"2025","journal-title":"IET Image Process"},{"key":"pone.0335615.ref037","article-title":"A survey of diffusion based image generation models: issues and their solutions.","author":"T Zhang","year":"2023","journal-title":"arXiv preprint"},{"issue":"12","key":"pone.0335615.ref038","doi-asserted-by":"crossref","DOI":"10.1093\/nsr\/nwae348","article-title":"Opportunities and challenges of diffusion models for generative AI","volume":"11","author":"M Chen","year":"2024","journal-title":"Natl Sci Rev"},{"key":"pone.0335615.ref039","doi-asserted-by":"crossref","first-page":"102846","DOI":"10.1016\/j.media.2023.102846","article-title":"Diffusion models in medical imaging: a comprehensive survey","volume":"88","author":"A Kazerouni","year":"2023","journal-title":"Med Image Anal"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0335615","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,29]],"date-time":"2025-10-29T17:33:33Z","timestamp":1761759213000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0335615"}},"subtitle":[],"editor":[{"given":"Andrzej","family":"Grzybowski","sequence":"first","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2025,10,29]]},"references-count":39,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2025,10,29]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0335615","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,10,29]]}}}