{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T05:51:44Z","timestamp":1791525104384,"version":"4.3.4"},"reference-count":29,"publisher":"IOP Publishing","issue":"19","license":[{"start":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T00:00:00Z","timestamp":1791504000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/publishingsupport.iopscience.iop.org\/iop-standard\/v1"},{"start":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T00:00:00Z","timestamp":1791504000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"name":"UAEU-AUA","award":["12R239"],"award-info":[{"award-number":["12R239"]}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Phys. Med. Biol."],"published-print":{"date-parts":[[2026,10,14]]},"abstract":"Abstract<\/jats:title>\n \n Objective.<\/jats:italic>\n Diffusion magnetic resonance imaging (MRI) signals are sensitive to microvascular geometry and hemodynamics, but current approaches rely on computationally intensive simulations on optical coherence tomography (OCT)-derived vascular graphs. We investigated whether three-dimensional deep learning models can serve as fast surrogates that predict diffusion MRI responses directly from microvascular OCT angiograms.\n Approach.<\/jats:italic>\n Five mouse cortical OCT angiograms were segmented into 4792 vascular subgraphs and converted into nine-channel image volumes encoding vascular geometry, magnetic field gradient, hematocrit, oxygen saturation, partial oxygen pressure, blood velocity magnitude, and three velocity components. For each sample, a physics-based framework generated spin echo (SE) and diffusion-weighted signals for three\n \n \n \n <\/jats:tex-math>\n \n \n b<\/mml:mi>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n -values (50, 100, 500 s\u2009mm\n \n \n \n <\/jats:tex-math>\n \n \n \n \n \n <\/mml:mrow>\n \n \u2212<\/mml:mo>\n 2<\/mml:mn>\n <\/mml:mrow>\n <\/mml:msup>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n ), three gradient timing pairs, and ten gradient directions. We trained five 3D architectures (AutoEncoder, BasicUNet, DenseNet169, EfficientNetB4 and ResNet) with late fusion of acquisition parameters and evaluated performance using MSE, MAE and\n \n \n \n <\/jats:tex-math>\n \n \n \n R<\/mml:mi>\n 2<\/mml:mn>\n <\/mml:msup>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n .\n Main results.<\/jats:italic>\n Under within-distribution (random-split) evaluation, test\n \n \n \n <\/jats:tex-math>\n \n \n \n R<\/mml:mi>\n 2<\/mml:mn>\n <\/mml:msup>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n values were at least 0.87. DenseNet169 achieved the best overall accuracy (MSE\n \n \n \n <\/jats:tex-math>\n \n \n =<\/mml:mo>\n 0.0039<\/mml:mn>\n \u00b1<\/mml:mo>\n 0.0029<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n , MAE\n \n \n \n <\/jats:tex-math>\n \n \n =<\/mml:mo>\n 0.0397<\/mml:mn>\n \u00b1<\/mml:mo>\n 0.0161<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n ,\n \n \n \n <\/jats:tex-math>\n \n \n \n R<\/mml:mi>\n 2<\/mml:mn>\n <\/mml:msup>\n =<\/mml:mo>\n 0.92<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n ) and maintained\n \n \n \n <\/jats:tex-math>\n \n \n \n R<\/mml:mi>\n 2<\/mml:mn>\n <\/mml:msup>\n ><\/mml:mo>\n 0.80<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n across all diffusion settings, while ResNet provided near-perfect reconstruction of the SE sequence alone (\n \n \n \n <\/jats:tex-math>\n \n \n \n R<\/mml:mi>\n 2<\/mml:mn>\n <\/mml:msup>\n \u2248<\/mml:mo>\n 1.00<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n ). Leave-one-subject-out cross-validation gave MSE\n \n \n \n <\/jats:tex-math>\n \n \n =<\/mml:mo>\n 0.0173<\/mml:mn>\n \u00b1<\/mml:mo>\n 0.0012<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n , MAE\n \n \n \n <\/jats:tex-math>\n \n \n =<\/mml:mo>\n 0.094<\/mml:mn>\n \u00b1<\/mml:mo>\n 0.004<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n , and\n \n \n \n <\/jats:tex-math>\n \n \n \n R<\/mml:mi>\n 2<\/mml:mn>\n <\/mml:msup>\n =<\/mml:mo>\n 0.666<\/mml:mn>\n \u00b1<\/mml:mo>\n 0.019<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n , with 34.4% of held-out samples reaching\n \n \n \n <\/jats:tex-math>\n \n \n \n R<\/mml:mi>\n 2<\/mml:mn>\n <\/mml:msup>\n \n \u2a7e<\/mml:mtext>\n <\/mml:mrow>\n 0.80<\/mml:mn>\n <\/mml:mrow>\n <\/mml:math>\n <\/jats:inline-formula>\n . Channel ablation revealed velocity magnitude and velocity components as dominant predictors. Once trained, the models predicted full diffusion responses in milliseconds, providing speedups of four to six orders of magnitude compared with simulations.\n Significance.<\/jats:italic>\n These results demonstrate that deep learning surrogates can accurately and efficiently approximate diffusion MRI responses from microvascular OCT angiograms, enabling rapid exploration of microvascular configurations and acquisition protocols.\n <\/jats:p>","DOI":"10.1088\/1361-6560\/aeaa9f","type":"journal-article","created":{"date-parts":[[2026,9,21]],"date-time":"2026-09-21T22:52:23Z","timestamp":1790031143000},"page":"195031","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Deep learning-based prediction of diffusion MRI responses from microvascular OCT angiograms"],"prefix":"10.1088","volume":"71","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3701-1611","authenticated-orcid":true,"given":"Jaloliddin","family":"Rustamov","sequence":"first","affiliation":[{"name":"United Arab Emirates University"},{"name":"ESAT\/PSI, KU Leuven"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"data-curation","vocabulary":"credit"},{"role":"formal-analysis","vocabulary":"credit"},{"role":"investigation","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"},{"role":"software","vocabulary":"credit"},{"role":"validation","vocabulary":"credit"},{"role":"visualization","vocabulary":"credit"},{"role":"writing-original-draft","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4977-1781","authenticated-orcid":true,"given":"Zahiriddin","family":"Rustamov","sequence":"additional","affiliation":[{"name":"United Arab Emirates University"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"data-curation","vocabulary":"credit"},{"role":"investigation","vocabulary":"credit"},{"role":"software","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3699-1283","authenticated-orcid":false,"given":"Fr\u00e9d\u00e9ric","family":"Lesage","sequence":"additional","affiliation":[{"name":"\u00c9cole Polytechnique de Montr\u00e9al"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"investigation","vocabulary":"credit"},{"role":"resources","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6259-9843","authenticated-orcid":false,"given":"Nazar","family":"Zaki","sequence":"additional","affiliation":[{"name":"United Arab Emirates University"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"writing-review-editing","vocabulary":"credit"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6797-0448","authenticated-orcid":false,"given":"Rafat","family":"Damseh","sequence":"additional","affiliation":[{"name":"United Arab Emirates University"}],"role":[{"vocabulary":"crossref","role":"author"},{"role":"conceptualization","vocabulary":"credit"},{"role":"funding-acquisition","vocabulary":"credit"},{"role":"methodology","vocabulary":"credit"},{"role":"project-administration","vocabulary":"credit"},{"role":"supervision","vocabulary":"credit"},{"role":"writing-review-editing","vocabulary":"credit"}]}],"member":"266","published-online":{"date-parts":[[2026,10,9]]},"reference":[{"key":"pmbaeaa9fbib1","article-title":"MONAI: an open-source framework for deep learning in healthcare","author":"Cardoso","year":"2022","type":"other"},{"key":"pmbaeaa9fbib2","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2022.102444","type":"journal-article","article-title":"Recent advances and clinical applications of deep learning in medical image analysis","volume":"79","author":"Chen","year":"2022","journal-title":"Med. Image Anal."},{"key":"pmbaeaa9fbib3","doi-asserted-by":"publisher","first-page":"381","DOI":"10.1109\/TMI.2020.3027500","type":"journal-article","article-title":"Laplacian flow dynamics on geometric graphs for anatomical modeling of cerebrovascular networks","volume":"40","author":"Damseh","year":"2020","journal-title":"IEEE Trans. Med. Imaging"},{"key":"pmbaeaa9fbib4","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-021-93503-2","type":"journal-article","article-title":"A simulation study investigating potential diffusion-based mri signatures of microstrokes","volume":"11","author":"Damseh","year":"2021","journal-title":"Sci. Rep."},{"key":"pmbaeaa9fbib5","doi-asserted-by":"publisher","first-page":"1","DOI":"10.5555\/1248547.1248548","type":"journal-article","article-title":"Statistical comparisons of classifiers over multiple data sets","volume":"7","author":"Dem\u0161ar","year":"2006","journal-title":"J. Mach. Learn. Res."},{"key":"pmbaeaa9fbib6","author":"Efron","year":"1994","type":"book"},{"key":"pmbaeaa9fbib7","article-title":"Performance analysis of unet and variants for medical image segmentation","author":"Ehab","year":"2023","type":"other"},{"key":"pmbaeaa9fbib8","first-page":"pp 3057","type":"conference-proceedings","article-title":"Data augmentation for deep learning based accelerated MRI reconstruction with limited data","volume":"vol 139","author":"Fabian","year":"2021"},{"key":"pmbaeaa9fbib9","doi-asserted-by":"publisher","first-page":"675","DOI":"10.1080\/01621459.1937.10503522","type":"journal-article","article-title":"The use of ranks to avoid the assumption of normality implicit in the analysis of variance","volume":"32","author":"Friedman","year":"1937","journal-title":"J. Am. Stat. Assoc."},{"key":"pmbaeaa9fbib10","doi-asserted-by":"publisher","first-page":"3663","DOI":"10.1523\/JNEUROSCI.3555-14.2015","type":"journal-article","article-title":"Quantifying the microvascular origin of bold-fMRI from first principles with two-photon microscopy and an oxygen-sensitive nanoprobe","volume":"35","author":"Gagnon","year":"2015","journal-title":"J. Neurosci."},{"key":"pmbaeaa9fbib11","doi-asserted-by":"publisher","first-page":"82","DOI":"10.3389\/fncom.2016.00082","type":"journal-article","article-title":"Modeling of cerebral oxygen transport based on in vivo microscopic imaging of microvascular network structure, blood flow and oxygenation","volume":"10","author":"Gagnon","year":"2016","journal-title":"Front. Comput. Neurosci."},{"key":"pmbaeaa9fbib12","doi-asserted-by":"publisher","first-page":"218","DOI":"10.1161\/STROKEAHA.118.022315","type":"journal-article","article-title":"Intravascular optical coherence tomography for neurointerventional surgery","volume":"50","author":"Gounis","year":"2019","journal-title":"Stroke"},{"key":"pmbaeaa9fbib13","doi-asserted-by":"publisher","first-page":"6953","DOI":"10.1109\/TIP.2025.3618390","type":"journal-article","article-title":"Physically consistent image augmentation for deep learning in Mueller matrix polarimetry","volume":"34","author":"Hahne","year":"2025","journal-title":"IEEE Trans. Image Process."},{"key":"pmbaeaa9fbib14","first-page":"pp 734","type":"conference-proceedings","article-title":"Gan-based synthetic brain MR image generation","author":"Han","year":"2018"},{"key":"pmbaeaa9fbib15","doi-asserted-by":"publisher","first-page":"2314","DOI":"10.3390\/app11052314","type":"journal-article","article-title":"Study on accuracy metrics for evaluating the predictions of damage locations in deep piles using artificial neural networks with acoustic emission data","volume":"11","author":"Jierula","year":"2021","journal-title":"Appl. Sci."},{"key":"pmbaeaa9fbib16","article-title":"Adam: a method for stochastic optimization","author":"Kingma","year":"2017","type":"other"},{"key":"pmbaeaa9fbib17","doi-asserted-by":"publisher","first-page":"56","DOI":"10.1016\/j.neuroimage.2017.12.062","type":"journal-article","article-title":"What can we see with IVIM MRI?","volume":"187","author":"Le Bihan","year":"2019","journal-title":"Neuroimage"},{"key":"pmbaeaa9fbib18","doi-asserted-by":"publisher","first-page":"1336","DOI":"10.1002\/mrm.21890","type":"journal-article","article-title":"The B-matrix must be rotated when correcting for subject motion in DTI data","volume":"61","author":"Leemans","year":"2009","journal-title":"Magn. Reson. Med."},{"key":"pmbaeaa9fbib19","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1016\/j.media.2017.07.005","type":"journal-article","article-title":"A survey on deep learning in medical image analysis","volume":"42","author":"Litjens","year":"2017","journal-title":"Med. Image Anal."},{"key":"pmbaeaa9fbib20","article-title":"Decoupled weight decay regularization","author":"Loshchilov","year":"2019","type":"conference-proceedings"},{"key":"pmbaeaa9fbib21","first-page":"p 4","type":"conference-proceedings","article-title":"Generative adversarial network based synthesis for supervised medical image segmentation","volume":"vol 3","author":"Neff","year":"2017"},{"key":"pmbaeaa9fbib22","article-title":"Distribution-free multiple comparisons","author":"Nemenyi","year":"1963","type":"other"},{"key":"pmbaeaa9fbib23","doi-asserted-by":"publisher","first-page":"2825","DOI":"10.5555\/1953048.2078195","type":"journal-article","article-title":"Scikit-learn: machine learning in python","volume":"12","author":"Pedregosa","year":"2011","journal-title":"J. Mach. Learn. Res."},{"key":"pmbaeaa9fbib24","doi-asserted-by":"publisher","first-page":"436","DOI":"10.1016\/j.neuroimage.2016.12.060","type":"journal-article","article-title":"Magnetic resonance fingerprinting based on realistic vasculature in mice","volume":"149","author":"Pouliot","year":"2017","journal-title":"Neuroimage"},{"key":"pmbaeaa9fbib25","first-page":"pp 54","type":"book","article-title":"Learning-based MRI response predictions from OCT microvascular models to replace simulation-based frameworks","author":"Rustamov","year":"2024"},{"key":"pmbaeaa9fbib26","type":"book","article-title":"Improved segmentation by adversarial U-Net","volume":"vol 11597","author":"Sriker","year":"2021"},{"key":"pmbaeaa9fbib27","doi-asserted-by":"publisher","first-page":"3851","DOI":"10.1038\/s41467-020-17702-7","type":"journal-article","article-title":"A neurovascular high-frequency optical coherence tomography system enables in situ cerebrovascular volumetric microscopy","volume":"11","author":"Ughi","year":"2020","journal-title":"Nat. Commun."},{"key":"pmbaeaa9fbib28","doi-asserted-by":"publisher","first-page":"137","DOI":"10.1007\/s12975-022-01023-6","type":"journal-article","article-title":"Optical coherence tomography in cerebrovascular disease: open up new horizons","volume":"14","author":"Xu","year":"2023","journal-title":"Transl. Stroke Res."},{"key":"pmbaeaa9fbib29","doi-asserted-by":"publisher","first-page":"3645","DOI":"10.1109\/TBME.2022.3175072","type":"journal-article","article-title":"Volumetric characterization of microvasculature in ex vivo human brain samples by serial sectioning optical coherence tomography","volume":"69","author":"Yang","year":"2022","journal-title":"IEEE Trans. Biomed. Eng."}],"container-title":["Physics in Medicine & Biology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,9]],"date-time":"2026-10-09T05:31:32Z","timestamp":1791523892000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1361-6560\/aeaa9f"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,9]]},"references-count":29,"journal-issue":{"issue":"19","published-online":{"date-parts":[[2026,10,9]]},"published-print":{"date-parts":[[2026,10,14]]}},"URL":"https:\/\/doi.org\/10.1088\/1361-6560\/aeaa9f","relation":{},"ISSN":["0031-9155","1361-6560"],"issn-type":[{"value":"0031-9155","type":"print"},{"value":"1361-6560","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,9]]},"assertion":[{"value":"Deep learning-based prediction of diffusion MRI responses from microvascular OCT angiograms","name":"article_title","label":"Article Title"},{"value":"Physics in Medicine & Biology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2026 Institute of Physics and Engineering in Medicine. All rights, including for text and data mining, AI training, and similar technologies, are reserved.","name":"copyright_information","label":"Copyright Information"},{"value":"2026-04-09","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-09-21","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-10-09","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}