CoChat

The DOI matches its Crossref record; 53 of the 55 references it cites were found in a registry or an index; 9 of the 22 sources its claims cite were checked for support.

Alzheimer's Disease and Frontotemporal Dementia: A Robust Classification Method of EEG Signals and a Comparison of Validation Methods

Open access
Miltiadous, Andreas, Tzimourta, Katerina D., Giannakeas, Nikolaos, Tsipouras, Markos G., Afrantou, Theodora, Ioannidis, Panagiotis, Tzallas, Alexandros T. · 2021 · Diagnostics

Abstract

Dementia is the clinical syndrome characterized by progressive loss of cognitive and emotional abilities to a degree severe enough to interfere with daily functioning. Alzheimer's disease (AD) is the most common neurogenerative disorder, making up 50-70% of total dementia cases. Another dementia type is frontotemporal dementia (FTD), which is associated with circumscribed degeneration of the prefrontal and anterior temporal cortex and mainly affects personality and social skills. With the rapid advancement in electroencephalogram (EEG) sensors, the EEG has become a suitable, accurate, and highly sensitive biomarker for the identification of neuronal and cognitive dynamics in most cases of dementia, such as AD and FTD, through EEG signal analysis and processing techniques. In this study, six supervised machine-learning techniques were compared on categorizing processed EEG signals of AD and FTD cases, to provide an insight for future methods on early dementia diagnosis. K-fold cross validation and leave-one-patient-out cross validation were also compared as validation methods to evaluate their performance for this classification problem. The proposed methodology accuracy scores were 78.5% for AD detection with decision trees and 86.3% for FTD detection with random forests.

Overview

CoChat Summary

## TL;DR

This paper compares six supervised classifiers on EEG features to distinguish AD, FTD, and healthy controls, and also compares 10-fold cross validation with leave-one-patient-out cross validation. The authors conclude that leave-one-patient-out is the more trustworthy validation method, with the best reported accuracies of 78.5% for AD/CN using decision trees and 86.3% for FTD/CN using random forests.

## Key findings

  • The study used six classifiers — decision trees, random forests, ANN, SVM, Naïve Bayes, and kNN — on three problems: AD/CN, FTD/CN, and AD/FTD.
  • Under leave-one-person-out cross validation, the best results were:
  • AD/CN: decision trees, 78.5% accuracy, 82.4% sensitivity, 74% specificity, SD 5.8.
  • FTD/CN: random forests, 86.3% accuracy, 87% sensitivity, 83% specificity, SD 7.1.
  • AD/FTD: decision trees, 73% accuracy, SD 11, with all algorithms performing poorly on this task.
  • Under 10-fold cross validation, results were “above 90%” for almost all classification problems and algorithms, but the authors argue this was inflated because epochs from the same subject appeared in both training and test sets.
  • The dataset consisted of EEG recordings from 28 participants: 10 AD, 10 FTD, and 8 healthy controls.

## Methods

  • Design: EEG resting-state recordings were preprocessed, features were extracted, and six supervised machine-learning algorithms were compared across three classification tasks (AD/CN, FTD/CN, AD/FTD). Validation methods compared were 10-fold cross validation and leave-one-person-out cross validation.
  • Data/sample: EEG recordings came from 28 participants at AHEPA General University Hospital of Thessaloniki: 10 AD, 10 FTD, and 8 age-matched controls. Recordings used 19 scalp electrodes in the 10–20 system.
  • Main analysis: Signals were artifact-cleaned, down-sampled from 500 Hz to 250 Hz, band-pass filtered at 0.5–48 Hz, segmented into 5 s epochs with 2.5 s overlap, and summarized with time-domain features (mean, variance, IQR) and frequency-domain band energies across delta, theta, alpha, beta, and gamma. Classification was implemented in Weka.

## Limitations

  • The paper explicitly states the clinical dataset was “relatively small,” with “18 or 20 subjects for each classification problem,” and says no clear conclusions can be drawn about generality from such a sample.
  • The authors note that leave-one-out validation has higher computational cost as dataset size increases, making it less appropriate for very large datasets.
  • Evident from the text: the AD/FTD classification performed poorly and had high standard deviations, and the authors suggest the preprocessing/classification setup may not suit that task or that the sample was too small.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/diagnostics11081437

Cites

53 of 53 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 OpenViBE Xml diagram of signal processing and feature extraction.
Figure 1 OpenViBE Xml diagram of signal processing and feature extraction.CC BY · PubMed Central
Figure 2 TP, TN, FP, FN representation.
Figure 2 TP, TN, FP, FN representation.CC BY · PubMed Central

Verification

Claims the paper makes

5 Valid · 1 Not valid · 5 Not checked · 25 Own results and interpretations

14 more claims

Create a free account to see all 36 claims and to verify the 1 reference still open.

See all claims — free account
  • Probably not a substantial claim. 8
  • It may state two separate claims as one. 11
  • The quote only partly states this claim. 3

Create a free account to see these claims.

Citations

Cites

Show all 53 references

Cited by

See all 162

Metadata

Authors
Miltiadous, A., Tzimourta, K. D., Giannakeas, N., Tsipouras, M. G., Afrantou, T., Ioannidis, P., & Tzallas, A. T.
Affiliations
  • Department of Informatics and Telecommunications, School of Informatics and Telecommunications, University of Ioannina, Kostakioi, 47 100 Arta, Greece
  • University of Western Macedonia (GR)
  • Aristotle University of Thessaloniki (GR)
  • University of Ioannina (GR)
Corresponding author
Tzallas, Alexandros T.
Venue
Diagnostics
Publisher
MDPI AG
Published
2021-08-09
Volume
11
Issue
8
Pages
1437
ISSN
2075-4418 (ISSN-L)
OpenAlex
W3189791507
Cited by
160
References
55
Citations by year
2026: 24 · 2025: 46 · 2024: 45 · 2023: 30 · 2022: 13 · 2021: 2
FWCI
11.83
Type
journal article
Language
en
Open access
Locations
  • Diagnostics (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Funders
  • European Commission (NSRF 2014-2020, 2014‐2020)
  • European Regional Development Fund (NSRF 2014–2020, 2014-2020’, MIS 5047221)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
  • Physical Sciences › Computer Science › Artificial Intelligence › Emotion and Mood Recognition
Keywords
Alzheimer's disease, frontotemporal dementia, EEG classification, dementia diagnosis, EEG biomarkers, supervised learning, decision tree, random forest, k-fold cross-validation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Alzheimer's Disease and Frontotemporal Dementia: A Robust Classification Method of EEG Signals and a Comparison of Validation Methods. https://app.cochat.ai/papers/10.3390/diagnostics11081437
The paper
Miltiadous, A., Tzimourta, K. D., Giannakeas, N., Tsipouras, M. G., Afrantou, T., Ioannidis, P., & Tzallas, A. T. (2021). Alzheimer's Disease and Frontotemporal Dementia: A Robust Classification Method of EEG Signals and a Comparison of Validation Methods. Diagnostics. https://doi.org/10.3390/diagnostics11081437

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.