{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"bioRxiv"}],"indexed":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T12:16:35Z","timestamp":1787919395777,"version":"build-2784847793"},"posted":{"date-parts":[[2026,8,24]]},"group-title":"Neuroscience","reference-count":18,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,8,24]],"date-time":"2026-08-24T00:00:00Z","timestamp":1787529600000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"National Institute on Aging, National Institutes of Health","award":["P30AG072958"],"award-info":[{"award-number":["P30AG072958"]}]},{"name":"National Institute on Aging, National Institutes of Health","award":["R01 AG066184"],"award-info":[{"award-number":["R01 AG066184"]}]},{"name":"National Institute on Aging, National Institutes of Health","award":["RF1 AG070149"],"award-info":[{"award-number":["RF1 AG070149"]}]},{"name":"National Institute on Aging, National Institutes of Health","award":["R01AG054073"],"award-info":[{"award-number":["R01AG054073"]}]},{"name":"National Institute on Aging, National Institutes of Health","award":["R01AG058533"],"award-info":[{"award-number":["R01AG058533"]}]},{"name":"National Institute on Aging, National Institutes of Health","award":["R01AG070862"],"award-info":[{"award-number":["R01AG070862"]}]},{"name":"National Institute of Biomedical Imaging and Bioengineering, National Institutes of Health","award":["P41EB015922"],"award-info":[{"award-number":["P41EB015922"]}]},{"name":"National Cancer Institute","award":["R01CA302738"],"award-info":[{"award-number":["R01CA302738"]}]},{"DOI":"10.13039\/100000050","name":"National Heart, Lung, and Blood Institute","doi-asserted-by":"crossref","award":["R01HL179758"],"award-info":[{"award-number":["R01HL179758"]}],"id":[{"id":"10.13039\/100000050","id-type":"DOI","asserted-by":"crossref"}]},{"id":[{"id":"https:\/\/ror.org\/01n9kga30","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,8,24]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                <jats:p>Brain-age models derived from diffusion MRI\u2013based structural connectomes may provide imaging biomarkers of accelerated brain aging, but their biological interpretation and transportability across heterogeneous populations remain uncertain. We developed a calibration-aware and hierarchically interpretable graph-learning framework and evaluated it across four independent aging and Alzheimer\u2019s disease\u2013related cohorts: ADNI, Duke\/UNC ADRC, HABS-HD, and AD-DECODE. The analysis included 1,093 connectome sessions from 789 participants. Cohort-specific graph neural networks were trained using participant-grouped cross-validation across five imaging and multimodal feature configurations. Prediction performance varied more strongly across cohorts than across feature sets, with imaging-only out-of-fold mean absolute error ranging from 4.72 years in ADNI to 9.75 years in AD-DECODE. The imaging-only graph neural network was competitive with ridge, elasticnet, and gradient-boosted regression models trained on matched vectorized connectome features, but was not uniformly superior. Age-bias-corrected brain-age gap was most consistently associated with reduced diffusion-derived microstructural integrity and structural-network organization across cohorts. In longitudinal analyses, corrected brain-age gap showed moderate-to-good within-person preservation in ADNI and HABS-HD, with intraclass correlation coefficients of 0.67 and 0.81, respectively; higher baseline values also predicted subsequent microstructural and network deterioration in ADNI. Multiscale SHAP analysis identified distributed contributions from global graph topology, regional imaging features, edge-derived regional summaries, and individual structural connections involving thalamic, striatal, frontal, parietal, cerebellar, hippocampal, and entorhinal circuitry. External transfer was highly sensitive to cohort shift: across 12 off-diagonal train\u2013test evaluations, median mean absolute error decreased from 17.39 to 8.22 years after target-cohort linear recalibration, whereas median Pearson correlation remained 0.17. Because recalibration used target-cohort chronological age, it was interpreted as a diagnostic sensitivity analysis rather than deployable external validation. Together, these findings support calibration-aware diffusion-connectome brain age as an interpretable imaging biomarker of structural brain aging and prospective microstructural and network vulnerability, while emphasizing the need for cohort-specific calibration before external application.<\/jats:p>","DOI":"10.64898\/2026.08.19.745783","type":"posted-content","created":{"date-parts":[[2026,8,24]],"date-time":"2026-08-24T21:01:21Z","timestamp":1787605281000},"source":"Crossref","is-referenced-by-count":0,"title":["Calibration-Aware and Interpretable Graph Learning for Multi-Cohort Diffusion Connectome Brain-Age Modeling"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6621-4560","authenticated-orcid":false,"given":"Alexandra","family":"Badea","sequence":"first","affiliation":[{"name":"Quantitative Imaging and Analysis Laboratories, Duke Radiology, Duke University, Durham, NC, USA"},{"name":"Neurology Department, Duke, University Medical Center, Durham, NC, USA"},{"name":"Brain Imaging and Analysis Center, Duke University Medical Center, Durham, NC, USA"},{"name":"Biomedical Engineering Department, Duke University, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ines Poves","family":"Acle","sequence":"additional","affiliation":[{"name":"Quantitative Imaging and Analysis Laboratories, Duke Radiology, Duke University, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Paula","family":"Mendez de Inza","sequence":"additional","affiliation":[{"name":"Quantitative Imaging and Analysis Laboratories, Duke Radiology, Duke University, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hanwen","family":"Lin","sequence":"additional","affiliation":[{"name":"Quantitative Imaging and Analysis Laboratories, Duke Radiology, Duke University, Durham, NC, USA"},{"name":"Biomedical Engineering Department, Duke University, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Robert J","family":"Anderson","sequence":"additional","affiliation":[{"name":"Quantitative Imaging and Analysis Laboratories, Duke Radiology, Duke University, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kim G","family":"Johnson","sequence":"additional","affiliation":[{"name":"Neurology Department, Duke, University Medical Center, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Heather E","family":"Whitson","sequence":"additional","affiliation":[{"name":"Duke Aging Center, Duke University Medical Center, Durham, NC, USA"},{"name":"Department of Medicine, Duke University Medical Center, Durham, NC, USA"},{"name":"Geriatrics Research Education and Clinical Center (GRECC), Durham VA Medical Center, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Allen W","family":"Song","sequence":"additional","affiliation":[{"name":"Brain Imaging and Analysis Center, Duke University Medical Center, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cristian T","family":"Badea","sequence":"additional","affiliation":[{"name":"Quantitative Imaging and Analysis Laboratories, Duke Radiology, Duke University, Durham, NC, USA"},{"name":"Biomedical Engineering Department, Duke University, Durham, NC, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"name":"the Alzheimer\u2019s Disease Neuroimaging Initiative","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"name":"the Health and Aging Brain Study (HABS-HD) Study Team","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026082804250496000_2026.08.19.745783v1.1","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2020.117441"},{"key":"2026082804250496000_2026.08.19.745783v1.2","doi-asserted-by":"crossref","unstructured":"Butler, Ellyn R. , Andrew Chen , Rabie Ramadan , Trang T. Le , Kosha Ruparel , Tyler M. Moore , Theodore D. Satterthwaite , Fengqing Zhang , Haochang Shou , and Ruben C. Gur . 2021. \u201cPitfalls in brain age analyses.\u201d In.: Wiley Online Library.","DOI":"10.1101\/2020.06.21.163741"},{"key":"2026082804250496000_2026.08.19.745783v1.3","first-page":"456","article-title":"Graph transformer geometric learning of brain networks using multimodal MR images for brain age estimation","volume":"42","year":"2022","journal-title":"IEEE Transactions on Medical Imaging"},{"key":"2026082804250496000_2026.08.19.745783v1.4","doi-asserted-by":"publisher","DOI":"10.1016\/j.neurobiolaging.2020.03.014"},{"key":"2026082804250496000_2026.08.19.745783v1.5","doi-asserted-by":"publisher","DOI":"10.1016\/j.tins.2017.10.001"},{"key":"2026082804250496000_2026.08.19.745783v1.6","unstructured":"Collins, Gary S. , Karel G. M. Moons , Paula Dhiman , Richard D. Riley , Andrew L. Beam , Ben van Calster , Marzyeh Ghassemi , Xiaoxuan Liu , Johannes B. Reitsma , and Maarten van Smeden . 2024. \u2019TRIPOD+ AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods\u2019, bmj."},{"key":"2026082804250496000_2026.08.19.745783v1.7","doi-asserted-by":"publisher","DOI":"10.1002\/hbm.25837"},{"key":"2026082804250496000_2026.08.19.745783v1.8","doi-asserted-by":"crossref","unstructured":"Gao, Jingjing , Jiaxin Liu , Yuhang Xu , Dawei Peng , and Zhengning Wang . 2023. \u2018Brain age prediction using the graph neural network based on resting-state functional MRI in Alzheimer\u2019s disease\u2019, Frontiers in Neuroscience, Volume 17\u20132023.","DOI":"10.3389\/fnins.2023.1222751"},{"key":"2026082804250496000_2026.08.19.745783v1.9","doi-asserted-by":"crossref","first-page":"744","DOI":"10.1038\/s43588-024-00659-8","article-title":"A perspective on brain-age estimation and its clinical promise","volume":"4","year":"2024","journal-title":"Nature computational science"},{"key":"2026082804250496000_2026.08.19.745783v1.10","doi-asserted-by":"crossref","unstructured":"Kawahara, Jeremy , Colin J. Brown , Steven P. Miller , Brian G. Booth , Vann Chau , Ruth E. Grunau , Jill G. Zwicker , and Ghassan Hamarneh . 2017. \u201cConvolutional neural networks for brain net-works; Towards predicting neurodevelopment.\u201d In.: Neuroimaging.","DOI":"10.1016\/j.neuroimage.2016.09.046"},{"key":"2026082804250496000_2026.08.19.745783v1.11","doi-asserted-by":"crossref","unstructured":"Kazi, Anees , Jocelyn Mora , Bruce Fischl , Adrian V. Dalca , and Iman Aganj . 2025. \u2019Structural Connectome Analysis using a Graph-based Deep Model for Prediction of Non-Imaging Variables\u2019, BioRxiv.","DOI":"10.1101\/2025.03.09.642165"},{"key":"2026082804250496000_2026.08.19.745783v1.12","doi-asserted-by":"publisher","DOI":"10.1093\/cercor\/bhaa066"},{"key":"2026082804250496000_2026.08.19.745783v1.13","unstructured":"Lundberg, Scott M. , and Su-In Lee . 2017. \u2019A unified approach to interpreting model predictions\u2019, Advances in neural information processing systems, 30."},{"key":"2026082804250496000_2026.08.19.745783v1.14","first-page":"e70140","article-title":"Health and Aging Brain Study\u2013 Health Disparities (HABS-HD) methods and partner characteristics\u2019","volume":"11","author":"O\u2019Bryant, and Habs- H. D. Study Team for the","year":"2025","journal-title":"Alzheimer\u2019s & Dementia: Translational Research & Clinical Interventions"},{"key":"2026082804250496000_2026.08.19.745783v1.15","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.2015142202"},{"key":"2026082804250496000_2026.08.19.745783v1.16","first-page":"4793","article-title":"A survey on explainable artificial intelligence (xai): Toward medical xai","volume":"32","year":"2020","journal-title":"IEEE transactions on neural networks and learning systems"},{"key":"2026082804250496000_2026.08.19.745783v1.17","doi-asserted-by":"crossref","first-page":"230","DOI":"10.1186\/s12916-019-1466-7","article-title":"Calibration: the Achilles heel of predictive analytics","volume":"17","year":"2019","journal-title":"BMC medicine"},{"key":"2026082804250496000_2026.08.19.745783v1.18","doi-asserted-by":"publisher","DOI":"10.1016\/j.jclinepi.2015.12.005"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.08.19.745783","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T11:25:15Z","timestamp":1787916315000},"score":1,"resource":{"primary":{"URL":"http:\/\/biorxiv.org\/lookup\/doi\/10.64898\/2026.08.19.745783"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,24]]},"references-count":18,"URL":"https:\/\/doi.org\/10.64898\/2026.08.19.745783","relation":{},"subject":[],"published":{"date-parts":[[2026,8,24]]},"subtype":"preprint"}}