{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,27]],"date-time":"2026-08-27T05:18:07Z","timestamp":1787807887082,"version":"build-2784847793"},"reference-count":35,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,27]],"date-time":"2026-08-27T00:00:00Z","timestamp":1787788800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Neuroinform."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>The differential diagnosis between Alzheimer\u2019s disease (AD) and frontotemporal dementia (FTD) presents a significant clinical challenge due to overlapping early-stage symptom profiles. Conventional resting-state EEG provides limited sensitivity to the impaired neural plasticity and lateralized cortical degeneration frequently observed in FTD.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We developed a domain-informed heterogeneous ensemble framework incorporating dynamic neural reactivity and hemispheric asymmetry metrics from 19-channel EEG recordings acquired from 88 participants (36 AD, 23 FTD, 29 cognitively normal controls) during resting-state and photic stimulation paradigms. A neural reactivity vector (V_diff) was derived to quantify state-dependent spectral transitions. The 1,014-dimensional feature space was reduced to 200 features via recursive feature elimination, prioritizing spectral power distributions, hemispheric asymmetry indices (HAI), and stimulation-induced reactivity parameters. A weighted ensemble of Extreme Gradient Boosting (XGBoost) and Random Forest classifiers was evaluated using a subject-aware 90\/10 holdout split with internal five-fold cross-validation.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The optimized model achieved a multi-class segment-level accuracy of 95.63% on an independently held-out test partition of 1,281 segments, with an internal five-fold cross-validation mean of 0.9846 \u00b1 0.003. FTD-specific precision reached 0.9907. SHAP analysis identified beta-band hemispheric asymmetry and alpha-band reactivity as the principal contributors to class separation.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>These findings indicate that the integration of dynamic state-transition measures with structural asymmetry proxies enhances electrophysiological discrimination between dementia subtypes. The framework provides a computationally efficient and biologically interpretable alternative to deep learning\u2013based methodologies for EEG-driven dementia classification.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fninf.2026.1902549","type":"journal-article","created":{"date-parts":[[2026,8,27]],"date-time":"2026-08-27T04:26:05Z","timestamp":1787804765000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Interpretable AI for neural signal decoding in dementia: an EEG ensemble approach to differential diagnosis"],"prefix":"10.3389","volume":"20","author":[{"given":"Fawad","family":"Muhammad","sequence":"first","affiliation":[{"name":"Department of Biomedical Engineering, Salim Habib University","place":["Karachi, Pakistan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Irfan Ahmed","family":"Usmani","sequence":"additional","affiliation":[{"name":"Department of Biomedical Engineering, Salim Habib University","place":["Karachi, Pakistan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Muhammad","family":"Aamir","sequence":"additional","affiliation":[{"name":"Department of Biomedical Engineering, Salim Habib University","place":["Karachi, Pakistan"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mai","family":"Alduailij","sequence":"additional","affiliation":[{"name":"Department of Computer Sciences, College of Computer and Information Sciences, Princess Nourah Bint Abdulrahman University","place":["Riyadh, Saudi Arabia"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mehrez","family":"Marzougui","sequence":"additional","affiliation":[{"name":"College of Computer Science, King Khalid University","place":["Abha, Saudi Arabia"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rab","family":"Nawaz","sequence":"additional","affiliation":[{"name":"School of Computer Science and Electronic Engineering (CSEE), University of Essex","place":["Colchester, United Kingdom"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,27]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","DOI":"10.1109\/JBHI.2025.3604638","article-title":"An AI-Driven interpretable multiview feature learning approach for EEG based epileptic seizure detection.","author":"Ahmad","year":"2025","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"B2","doi-asserted-by":"publisher","first-page":"1270405","DOI":"10.3389\/fneur.2023.1270405","article-title":"Using shallow neural networks with functional connectivity from EEG signals for early diagnosis of Alzheimer\u2019s and frontotemporal dementia.","volume":"14","author":"Ajra","year":"2023","journal-title":"Front. Neurol."},{"key":"B3","doi-asserted-by":"publisher","first-page":"7","DOI":"10.3390\/engproc2023050007","article-title":"Brain signals to actions using machine learning.","volume":"50","author":"Angelakis","year":"2023","journal-title":"Eng. Proc."},{"key":"B4","doi-asserted-by":"publisher","first-page":"314","DOI":"10.3389\/fnagi.2014.00314","article-title":"Slowing of EEG background activity in Parkinson\u00e2\u20ac\u2122s and Alzheimer\u00e2\u20ac\u2122s disease with early cognitive dysfunction.","volume":"6","author":"Benz","year":"2014","journal-title":"Front. Aging Neurosci."},{"key":"B5","doi-asserted-by":"crossref","DOI":"10.1109\/BIBE66822.2025.00107","article-title":"NeuroXAI: Explainable Deep Learning for EEG-Based Detection of Alzheimer\u2019s and Parkinson\u2019s Diseases","volume-title":"Proceedings of the 2025 IEEE 25th International Conference on Bioinformatics and Bioengineering (BIBE)","author":"Bunterngchit","year":"2025"},{"key":"B6","doi-asserted-by":"crossref","DOI":"10.1109\/REM63063.2024.10735597","article-title":"Advanced EEG-Based Classification of Alzheimer\u2019s Disease Using CNN-LSTM-Attention Architecture","volume-title":"Proceedings of the 2024 22nd International Conference on Research and Education in Mechatronics (REM)","author":"Bunterngchit","year":"2024"},{"key":"B7","doi-asserted-by":"publisher","first-page":"111076","DOI":"10.1016\/j.compeleceng.2026.111076","article-title":"Attention-guided electroencephalography fusion with fractal and nonlinear dynamics decodes early Alzheimer\u2019s biomarkers.","volume":"133","author":"Bunterngchit","year":"2026","journal-title":"Comput. Electr. Eng."},{"key":"B8","doi-asserted-by":"publisher","first-page":"106778","DOI":"10.1016\/j.neunet.2024.106778","article-title":"A deep learning approach for non-invasive Alzheimer\u2019s monitoring using microwave radar data.","volume":"181","author":"Farhatullah, Chen","year":"2025","journal-title":"Neural Netw."},{"key":"B9","doi-asserted-by":"publisher","first-page":"104","DOI":"10.1186\/s12910-024-01103-2","article-title":"The ethical requirement of explainability for AI-DSS in healthcare: A systematic review of reasons.","volume":"25","author":"Freyer","year":"2024","journal-title":"BMC Med. Ethics"},{"key":"B10","doi-asserted-by":"publisher","first-page":"55","DOI":"10.1109\/TIT.1968.1054102","article-title":"On the mean accuracy of statistical pattern recognizers.","volume":"14","author":"Hughes","year":"1968","journal-title":"IEEE Trans. Inf. Theory"},{"key":"B11","article-title":"Combination of Bagging and Boosting Ensemble Learning Methods: A Solution to Model Underfitting and Overfitting","volume-title":"Proceedings of the LASUSTECH 30th iSTEAMS Multidisciplinary Innovations Conference","author":"Johnson","year":"2022"},{"key":"B12","doi-asserted-by":"publisher","first-page":"1706099","DOI":"10.3389\/fninf.2025.1706099","article-title":"CNN-based framework for Alzheimer\u2019s disease detection from EEG via dynamic mode decomposition.","volume":"19","author":"Kang","year":"2025","journal-title":"Front. Neuroinform."},{"key":"B13","doi-asserted-by":"publisher","first-page":"26","DOI":"10.3390\/a13010026","article-title":"Soft-Voting ensemble based co-training scheme using static selection for binary classification problems.","volume":"13","author":"Karlos","year":"2020","journal-title":"Algorithms"},{"key":"B14","doi-asserted-by":"publisher","first-page":"1590201","DOI":"10.3389\/fmed.2025.1590201","article-title":"An explainable and efficient deep learning framework for EEG-based diagnosis of Alzheimer\u2019s disease and frontotemporal dementia.","volume":"12","author":"Khan","year":"2025","journal-title":"Front. Med."},{"key":"B15","doi-asserted-by":"publisher","first-page":"73","DOI":"10.1016\/j.patcog.2015.11.015","article-title":"Overfitting in linear feature extraction for classification of high-dimensional image data.","volume":"53","author":"Liu","year":"2016","journal-title":"Pattern Recogn."},{"key":"B16","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1706.06060","article-title":"Consistent feature attribution for tree ensembles.","author":"Lundberg","year":"2018","journal-title":"arXiv [Preprint]"},{"key":"B17","doi-asserted-by":"publisher","first-page":"105431","DOI":"10.1016\/j.bspc.2023.105431","article-title":"Attention-related power and functional connectivity modulation associated with long-term alpha neurofeedback training.","volume":"87","author":"Mahmood","year":"2024","journal-title":"Biomed. Signal Process Control"},{"key":"B18","doi-asserted-by":"publisher","first-page":"213","DOI":"10.1186\/1471-2105-10-213","article-title":"A comparison of random forest and its Gini importance with standard chemometric methods for the feature selection and classification of spectral data.","volume":"10","author":"Menze","year":"2009","journal-title":"BMC Bioinform."},{"key":"B19","doi-asserted-by":"publisher","first-page":"1437","DOI":"10.3390\/diagnostics11081437","article-title":"Alzheimer\u2019s disease and frontotemporal dementia: A robust classification method of EEG signals and a comparison of validation methods.","volume":"11","author":"Miltiadous","year":"2021","journal-title":"Diagnostics"},{"key":"B20","doi-asserted-by":"publisher","first-page":"4903","DOI":"10.1038\/s41598-026-35316-9","article-title":"EEG-based classification of alzheimer\u2019s disease and frontotemporal dementia using functional connectivity.","volume":"16","author":"Mlinari\u010d","year":"2026","journal-title":"Sci. Rep."},{"key":"B21","doi-asserted-by":"publisher","first-page":"e105","DOI":"10.1016\/S2468-2667(21)00249-8","article-title":"Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: An analysis for the Global Burden of Disease Study 2019.","volume":"7","author":"Nichols","year":"2022","journal-title":"Lancet Public Health"},{"key":"B22","doi-asserted-by":"publisher","first-page":"64","DOI":"10.3390\/data10050064","article-title":"A complementary dataset of scalp EEG recordings featuring participants with Alzheimer\u2019s disease, frontotemporal dementia, and healthy controls, obtained from photostimulation EEG.","volume":"10","author":"Ntetska","year":"2025","journal-title":"Data"},{"key":"B23","doi-asserted-by":"publisher","first-page":"169","DOI":"10.1613\/jair.614","article-title":"Popular ensemble methods: An empirical study.","volume":"11","author":"Opitz","year":"1999","journal-title":"J. Art. Intell. Res."},{"key":"B24","doi-asserted-by":"publisher","first-page":"4713","DOI":"10.1007\/s00330-025-11400-y","article-title":"The cortical asymmetry index for subtyping dementia patients.","volume":"35","author":"P\u00e9rez-Millan","year":"2025","journal-title":"Eur. Radiol."},{"key":"B25","doi-asserted-by":"publisher","first-page":"2277","DOI":"10.1212\/wnl.54.12.2277","article-title":"Differentiating frontal and temporal variant frontotemporal dementia from Alzheimer\u2019s disease.","volume":"54","author":"Perry","year":"2000","journal-title":"Neurology"},{"key":"B26","doi-asserted-by":"publisher","first-page":"3447","DOI":"10.1007\/s11571-024-10152-7","article-title":"EEG-based classification of Alzheimer\u2019s disease and frontotemporal dementia: A comprehensive analysis of discriminative features.","volume":"18","author":"Rostamikia","year":"2024","journal-title":"Cogn. Neurodyn."},{"key":"B27","doi-asserted-by":"publisher","first-page":"265","DOI":"10.1186\/s13195-025-01875-8","article-title":"EEG-based detection of early functional brain changes in subjective cognitive decline: A prospective cohort study.","volume":"17","author":"Ryoo","year":"2025","journal-title":"Alzheimers Res. Ther."},{"key":"B28","doi-asserted-by":"publisher","first-page":"241","DOI":"10.2478\/jccm-2021-0022","article-title":"Accuracy of diagnostic tests.","volume":"7","author":"Santini","year":"2021","journal-title":"J. Crit. Care Med."},{"key":"B29","doi-asserted-by":"publisher","first-page":"4521","DOI":"10.1109\/TNSRE.2023.3329174","article-title":"Differentiating between Alzheimer\u2019s disease and frontotemporal dementia based on the resting-state multilayer EEG network.","volume":"31","author":"Si","year":"2023","journal-title":"IEEE Trans. Neural Syst. Rehabil. Eng."},{"key":"B30","doi-asserted-by":"publisher","first-page":"1617883","DOI":"10.3389\/fncom.2025.1617883","article-title":"Privacy\u2013preserving dementia classification from EEG via hybrid\u2013fusion EEGNetv4 and federated learning.","volume":"19","author":"Umair","year":"2025","journal-title":"Front. Comput. Neurosci."},{"key":"B31","doi-asserted-by":"publisher","first-page":"6077","DOI":"10.3390\/s23136077","article-title":"k-Fold cross-validation can significantly over-estimate true classification accuracy in common EEG-Based passive BCI experimental designs: An empirical investigation.","volume":"23","author":"White","year":"2023","journal-title":"Sensors"},{"key":"B32","doi-asserted-by":"publisher","first-page":"e0281153","DOI":"10.1371\/journal.pone.0281153","article-title":"Estimating the lifetime risk of a false positive screening test result.","volume":"18","author":"White","year":"2023","journal-title":"PLoS One"},{"key":"B33","doi-asserted-by":"publisher","first-page":"427","DOI":"10.3390\/e23040427","article-title":"Brain dynamics altered by photic stimulation in patients with Alzheimer\u2019s disease and mild cognitive impairment.","volume":"23","author":"Yu","year":"2021","journal-title":"Entropy"},{"key":"B34","doi-asserted-by":"publisher","first-page":"16341","DOI":"10.18632\/aging.103676","article-title":"Abnormal static and dynamic functional connectivity of resting-state fMRI in multiple system atrophy.","volume":"12","author":"Zheng","year":"2020","journal-title":"Aging"},{"key":"B35","doi-asserted-by":"publisher","first-page":"109","DOI":"10.4137\/CIN.S30804","article-title":"High-Dimensional statistical learning: Roots, justifications, and potential machineries.","volume":"14","author":"Zollanvari","year":"2016","journal-title":"Cancer Inform."}],"container-title":["Frontiers in Neuroinformatics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fninf.2026.1902549\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,27]],"date-time":"2026-08-27T04:26:07Z","timestamp":1787804767000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fninf.2026.1902549\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,27]]},"references-count":35,"alternative-id":["10.3389\/fninf.2026.1902549"],"URL":"https:\/\/doi.org\/10.3389\/fninf.2026.1902549","relation":{},"ISSN":["1662-5196"],"issn-type":[{"value":"1662-5196","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,27]]},"article-number":"1902549"}}