{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T12:56:51Z","timestamp":1785589011431,"version":"3.56.0"},"posted":{"date-parts":[[2026,7,29]]},"group-title":"Neurology","reference-count":67,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,7,29]],"date-time":"2026-07-29T00:00:00Z","timestamp":1785283200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,7,29]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>EEG-based machine learning shows promise for neurodegenerative disease classification, but robustness to sample imbalance, center heterogeneity, and validation leakage remains a key concern for clinical translation. We developed a new framework to assess diagnostic performance, calibration, and cross-center generalizability of EEG multifeatured classifiers across CN (cognitively normal), MCI (mild cognitive impairment), AD (Alzheimer\u2019s disease), and FTD (frontotemporal dementia), while addressing imbalance, statistical uncertainty, and validation rigor across six centers. Supervised classifiers were evaluated at aggregated- and subject-level repeated cross-validation and leave-one-center-out (LOCO) schemes, and calibration was implemented via Platt scaling within strictly nested folds. CN vs AD classification showed robust performance and cross-center generalizability, with consistent AUC and calibration across cross-validation and leave-one-center-out analyses. In contrast, CN versus MCI showed moderate, heterogeneous performance and limited cross-center generalizability, with chance-level results in some cohorts, while MCI versus AD showed moderate discrimination in a single available center. FTD contrasts showed modest or limited performance due to sparse samples. Predicted probabilities were stable across validation regimes for AD, but less consistent for MCI and FTD, and correlated robustly with cognitive impairment severity only for AD. Feature importance analyses identified disease-specific signatures, including alpha-band degradation and slow-wave increases in AD, with weaker and more heterogeneous patterns in prodromal and differential dementia contrasts (FTD vs AD). EEG classifiers provided robust discrimination for CN vs AD but showed limited and heterogeneous performance for MCI and FTD across centers. These results emphasize the need for balanced sampling, strict validation of clinical and EEG protocols, and uncertainty quantification to support reliable clinical deployment.<\/jats:p>","DOI":"10.64898\/2026.07.28.26359133","type":"posted-content","created":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T01:40:11Z","timestamp":1785375611000},"source":"Crossref","is-referenced-by-count":0,"title":["Generalizability of EEG-Based Dementia Classifiers: A Multicenter study of Alzheimer\u2019s, MCI, And FTD"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4093-7135","authenticated-orcid":false,"given":"Laouen","family":"Belloli","sequence":"first","affiliation":[{"name":"Institut du Cerveau, ICM, Inserm, CNRS, Sorbonne Universit\u00e9, Paris, France"},{"name":"ICC, CONICET & Universidad de Buenos Aires, Argentina"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8822-8146","authenticated-orcid":false,"given":"Nicol\u00e1s","family":"Bruno","sequence":"additional","affiliation":[{"name":"Institut du Cerveau, ICM, Inserm, CNRS, Sorbonne Universit\u00e9, Paris, France"},{"name":"Instituto de F\u00edsica Aplicada, CONICET & Universidad de Buenos Aires, Argentina"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hernan","family":"Hernandez","sequence":"additional","affiliation":[{"name":"Latin American Brain Health Institute (BrainLat), Universidad Adolfo Ib\u00e1\u00f1ez, Santiago, Chile"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jhosmary","family":"Cuadros","sequence":"additional","affiliation":[{"name":"Latin American Brain Health Institute (BrainLat), Universidad Adolfo Ib\u00e1\u00f1ez, Santiago, Chile"},{"name":"Advanced Center for Electrical and Electronic Engineering, Universidad T\u00e9cnica Federico Santa Mar\u00eda, Valpara\u00edso 2390123, Chile"},{"name":"Grupo de Bioingenier\u00eda, Decanato de Investigaci\u00f3n, Universidad Nacional Experimental del T\u00e1chira, San Crist\u00f3bal 5001, Venezuela"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0472-8472","authenticated-orcid":false,"given":"Dami\u00e1n","family":"Dellavale","sequence":"additional","affiliation":[{"name":"Global Brain Health Institute (GBHI), Trinity College Dublin (TCD), Dublin, Ireland"},{"name":"Consejo Nacional de Investigaciones Cient\u00edficas y T\u00e9cnicas (CONICET), Centro At\u00f3mico Bariloche, San Carlos de Bariloche, R\u00edo Negro, Argentina"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Pavel","family":"Prado","sequence":"additional","affiliation":[{"name":"Escuela de Fonoaudiolog\u00eda, Facultad de Ciencias de la Rehabilitaci\u00f3n y Calidad de Vida, Universidad San Sebasti\u00e1n, Santiago de Chile, Chile"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Renato","family":"Anghinah","sequence":"additional","affiliation":[{"name":"Cognition Rehabilitation after TBI Center Universidade de Sao Paulo, S\u00e3o Paulo, Brazil"},{"name":"Department of Neurology, Faculdade de Medicina do ABC, Santo Andr\u00e9, Brazil"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bahar","family":"G\u00fcntekin","sequence":"additional","affiliation":[{"name":"Istanbul Medipol University, School of Medicine, Department of Biophysics, T\u00fcrkiye"},{"name":"Brain and Cognition Research Center, BEYKOG, \u1fd9stanbul Medipol University, T\u00fcrkiye"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"L\u00fctf\u00fc","family":"Hano\u01e7lu","sequence":"additional","affiliation":[{"name":"Brain and Cognition Research Center, BEYKOG, \u1fd9stanbul Medipol University, T\u00fcrkiye"},{"name":"Istanbul Medipol University, School of Medicine, Department of Neurology, T\u00fcrkiye"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mario A","family":"Parra","sequence":"additional","affiliation":[{"name":"Department of Psychological Sciences and Health, University of Strathclyde, Glasgow, UK"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6758-5101","authenticated-orcid":false,"given":"Agust\u00edn","family":"Iba\u00f1ez","sequence":"additional","affiliation":[{"name":"Instituto de F\u00edsica Aplicada, CONICET & Universidad de Buenos Aires, Argentina"},{"name":"Latin American Brain Health Institute (BrainLat), Universidad Adolfo Ib\u00e1\u00f1ez, Santiago, Chile"},{"name":"Global Brain Health institute, University of California San francisco, San francisco, USA"},{"name":"Global Brain Health Institute (GBHI), Trinity College Dublin (TCD), Dublin, Ireland"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3878-4846","authenticated-orcid":false,"given":"Jacobo","family":"Sitt","sequence":"additional","affiliation":[{"name":"Institut du Cerveau, ICM, Inserm, CNRS, Sorbonne Universit\u00e9, Paris, France"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026080105000483000_2026.07.28.26359133v1.1","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-67448-3"},{"key":"2026080105000483000_2026.07.28.26359133v1.2","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroscience.2025.02.016"},{"key":"2026080105000483000_2026.07.28.26359133v1.3","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-025-03808-2"},{"key":"2026080105000483000_2026.07.28.26359133v1.4","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-024-03209-x"},{"key":"2026080105000483000_2026.07.28.26359133v1.5","doi-asserted-by":"publisher","DOI":"10.1002\/dad2.12455"},{"key":"2026080105000483000_2026.07.28.26359133v1.6","doi-asserted-by":"publisher","DOI":"10.1038\/s44220-025-00502-7"},{"key":"2026080105000483000_2026.07.28.26359133v1.7","doi-asserted-by":"publisher","DOI":"10.1016\/S2468-2667(21)00249-8"},{"key":"2026080105000483000_2026.07.28.26359133v1.8","doi-asserted-by":"publisher","DOI":"10.3389\/fnagi.2017.00221"},{"key":"2026080105000483000_2026.07.28.26359133v1.9","doi-asserted-by":"publisher","DOI":"10.1016\/S0140-6736(20)30367-6"},{"key":"2026080105000483000_2026.07.28.26359133v1.10","doi-asserted-by":"publisher","DOI":"10.1038\/s41597-023-02806-8"},{"key":"2026080105000483000_2026.07.28.26359133v1.11","doi-asserted-by":"publisher","DOI":"10.1002\/bsa3.70027"},{"key":"2026080105000483000_2026.07.28.26359133v1.12","doi-asserted-by":"publisher","DOI":"10.1016\/j.jalz.2018.02.018"},{"key":"2026080105000483000_2026.07.28.26359133v1.13","doi-asserted-by":"publisher","DOI":"10.1016\/j.jalz.2011.03.005"},{"key":"2026080105000483000_2026.07.28.26359133v1.14","doi-asserted-by":"publisher","DOI":"10.1016\/S1474-4422(14)70090-0"},{"key":"2026080105000483000_2026.07.28.26359133v1.15","doi-asserted-by":"publisher","DOI":"10.1002\/alz.12202"},{"issue":"2","key":"2026080105000483000_2026.07.28.26359133v1.16","doi-asserted-by":"crossref","first-page":"721","DOI":"10.1002\/alz.12757","article-title":"Biomarkers for dementia in latin american countries: Gaps and opportunities","volume":"19","year":"2023","journal-title":"Alzheimer\u2019s & Dementia"},{"key":"2026080105000483000_2026.07.28.26359133v1.17","doi-asserted-by":"publisher","DOI":"10.1002\/alz.14289"},{"key":"2026080105000483000_2026.07.28.26359133v1.18","doi-asserted-by":"publisher","DOI":"10.3389\/fneur.2021.663407"},{"key":"2026080105000483000_2026.07.28.26359133v1.19","doi-asserted-by":"publisher","DOI":"10.3390\/brainsci11020244"},{"key":"2026080105000483000_2026.07.28.26359133v1.20","doi-asserted-by":"publisher","DOI":"10.1212\/WNL.0000000000004897"},{"key":"2026080105000483000_2026.07.28.26359133v1.21","doi-asserted-by":"publisher","DOI":"10.1002\/alz.13788"},{"key":"2026080105000483000_2026.07.28.26359133v1.22","doi-asserted-by":"publisher","DOI":"10.1038\/s43587-025-01061-3"},{"key":"2026080105000483000_2026.07.28.26359133v1.23","doi-asserted-by":"publisher","DOI":"10.1016\/j.clinph.2020.03.003"},{"key":"2026080105000483000_2026.07.28.26359133v1.24","doi-asserted-by":"publisher","DOI":"10.1038\/s41398-022-02237-w"},{"key":"2026080105000483000_2026.07.28.26359133v1.25","doi-asserted-by":"publisher","DOI":"10.1002\/alz.12311"},{"key":"2026080105000483000_2026.07.28.26359133v1.26","doi-asserted-by":"publisher","DOI":"10.4061\/2011\/539621"},{"key":"2026080105000483000_2026.07.28.26359133v1.27","doi-asserted-by":"publisher","DOI":"10.1016\/S0197-4580(00)00153-6"},{"key":"2026080105000483000_2026.07.28.26359133v1.28","doi-asserted-by":"publisher","DOI":"10.1016\/j.clinph.2011.02.011"},{"key":"2026080105000483000_2026.07.28.26359133v1.29","doi-asserted-by":"publisher","DOI":"10.1016\/j.brainresbull.2025.111281"},{"key":"2026080105000483000_2026.07.28.26359133v1.30","doi-asserted-by":"publisher","DOI":"10.3389\/fnagi.2025.1522552"},{"key":"2026080105000483000_2026.07.28.26359133v1.31","doi-asserted-by":"publisher","DOI":"10.1088\/1741-2552\/ac87d0"},{"key":"2026080105000483000_2026.07.28.26359133v1.32","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijpsycho.2021.12.008"},{"key":"2026080105000483000_2026.07.28.26359133v1.33","doi-asserted-by":"publisher","DOI":"10.1142\/S0129065721500167"},{"key":"2026080105000483000_2026.07.28.26359133v1.34","doi-asserted-by":"publisher","DOI":"10.21203\/rs.3.rs-8071648\/v1"},{"key":"2026080105000483000_2026.07.28.26359133v1.35","doi-asserted-by":"publisher","DOI":"10.1007\/s11571-023-09993-5"},{"key":"2026080105000483000_2026.07.28.26359133v1.36","doi-asserted-by":"publisher","DOI":"10.1093\/brain\/awu141"},{"key":"2026080105000483000_2026.07.28.26359133v1.37","doi-asserted-by":"publisher","DOI":"10.1016\/j.cub.2013.07.075"},{"key":"2026080105000483000_2026.07.28.26359133v1.38","doi-asserted-by":"publisher","DOI":"10.1093\/brain\/awz150"},{"key":"2026080105000483000_2026.07.28.26359133v1.39","doi-asserted-by":"publisher","DOI":"10.1016\/j.neurobiolaging.2021.04.024"},{"key":"2026080105000483000_2026.07.28.26359133v1.40","doi-asserted-by":"publisher","DOI":"10.1093\/brain\/awy251"},{"key":"2026080105000483000_2026.07.28.26359133v1.41","doi-asserted-by":"publisher","DOI":"10.1093\/nc\/niae027"},{"key":"2026080105000483000_2026.07.28.26359133v1.42","doi-asserted-by":"publisher","DOI":"10.1101\/2024.11.22.24317805"},{"key":"2026080105000483000_2026.07.28.26359133v1.43","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2022.119003"},{"key":"2026080105000483000_2026.07.28.26359133v1.44","doi-asserted-by":"publisher","DOI":"10.1038\/s41593-023-01449-7"},{"key":"2026080105000483000_2026.07.28.26359133v1.45","doi-asserted-by":"publisher","DOI":"10.1145\/1102351"},{"key":"2026080105000483000_2026.07.28.26359133v1.46","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2013.10.027"},{"key":"2026080105000483000_2026.07.28.26359133v1.47","doi-asserted-by":"crossref","first-page":"035009","DOI":"10.1088\/1361-6579\/ac890d","article-title":"Quantitative signal quality assessment for large-scale continuous scalp electroencephalography from a big data perspective","volume":"44","year":"2023","journal-title":"Physiological Measurement"},{"key":"2026080105000483000_2026.07.28.26359133v1.48","doi-asserted-by":"publisher","DOI":"10.1007\/s40031-025-01259-4"},{"key":"2026080105000483000_2026.07.28.26359133v1.49","doi-asserted-by":"publisher","DOI":"10.1186\/s12916-019-1466-7"},{"key":"2026080105000483000_2026.07.28.26359133v1.50","doi-asserted-by":"publisher","DOI":"10.1101\/2025.09.10.25334048"},{"key":"2026080105000483000_2026.07.28.26359133v1.51","doi-asserted-by":"crossref","unstructured":"Peng, B. , Du, L. , Dang, M. , Li, T. , Li, Z. , Liu, J. , Chen, Y. , Liu, B. , & Zhang, Z. (2026). Decoupling mci-specific signatures from shared neurobiological substrates of cognitive aging via deep learning. npj Digital Medicine.","DOI":"10.1038\/s41746-026-02597-3"},{"key":"2026080105000483000_2026.07.28.26359133v1.52","doi-asserted-by":"publisher","DOI":"10.1186\/s13195-023-01181-1"},{"key":"2026080105000483000_2026.07.28.26359133v1.53","doi-asserted-by":"publisher","DOI":"10.3389\/fnagi.2013.00058"},{"key":"2026080105000483000_2026.07.28.26359133v1.54","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2020.116795"},{"key":"2026080105000483000_2026.07.28.26359133v1.55","doi-asserted-by":"publisher","DOI":"10.1016\/j.biopsych.2024.05.025"},{"key":"2026080105000483000_2026.07.28.26359133v1.56","doi-asserted-by":"publisher","DOI":"10.1111\/joim.12190"},{"key":"2026080105000483000_2026.07.28.26359133v1.57","doi-asserted-by":"publisher","DOI":"10.1016\/j.clinph.2025.2110931"},{"key":"2026080105000483000_2026.07.28.26359133v1.58","doi-asserted-by":"publisher","DOI":"10.1007\/s11571-024-10152-7"},{"key":"2026080105000483000_2026.07.28.26359133v1.59","doi-asserted-by":"publisher","DOI":"10.1002\/14651858.CD010783.pub2"},{"key":"2026080105000483000_2026.07.28.26359133v1.60","doi-asserted-by":"publisher","DOI":"10.1002\/alz.13472"},{"key":"2026080105000483000_2026.07.28.26359133v1.61","doi-asserted-by":"publisher","DOI":"10.1002\/gps.2314"},{"key":"2026080105000483000_2026.07.28.26359133v1.62","doi-asserted-by":"publisher","DOI":"10.1097\/WAD.0b013e318211c83c"},{"key":"2026080105000483000_2026.07.28.26359133v1.63","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.20206227"},{"key":"2026080105000483000_2026.07.28.26359133v1.64","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.20208432"},{"issue":"31","key":"2026080105000483000_2026.07.28.26359133v1.65","doi-asserted-by":"crossref","first-page":"1026","DOI":"10.21105\/joss.01026","article-title":"Pingouin: Statistics in python","volume":"3","year":"2018","journal-title":"J. Open Source Softw"},{"key":"2026080105000483000_2026.07.28.26359133v1.66","unstructured":"Lundberg, S. , & Lee, S.-I. (2017). A unified approach to interpreting model predictions. arXiv [cs.AI]."},{"key":"2026080105000483000_2026.07.28.26359133v1.67","doi-asserted-by":"publisher","DOI":"10.1038\/s42256-019-0138-9"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.07.28.26359133","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,1]],"date-time":"2026-08-01T12:00:14Z","timestamp":1785585614000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.07.28.26359133"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,29]]},"references-count":67,"URL":"https:\/\/doi.org\/10.64898\/2026.07.28.26359133","relation":{},"subject":[],"published":{"date-parts":[[2026,7,29]]},"subtype":"preprint"}}