EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Self-Organized Explainable Symbolic EEG Feature Engineering Using Directed Lobish and Previous-Next Pattern: A Proof-of-Concept Study on Dementia Classification

Self-Organized Explainable Symbolic EEG Feature Engineering Using Directed Lobish and Previous-Next Pattern: Eine Proof-of-Concept-Studie zur Demenzklassifikation

Ferhat Balgetir iD, Prabal Datta Barua iD, Jun Shen iD, Padmanesan Narasimhan iD, Smitha Shivshankar, Sengul Dogan iD, Irem Tasci, Burak Tasci iD, Mehmet Baygin iD, Turker Tuncer iD, Makiko Kobayashi iD

Journal of Clinical Medicine · 03.11.2026 · 15 (21) · 8572

DOI: 10.3390/jcm15218572 ↗

Auf einen Blick

Eine EEG-basierte, interpretierbare Merkmalskette wurde bei Alzheimer-Krankheit, frontotemporaler Demenz und Kontrollen getestet. Die Genauigkeit war bei segmentweiser Validierung sehr hoch, bei teilnehmerunabhängiger Validierung deutlich niedriger.

Fragestellung

Wie leistungsfähig ist ein interpretierbares symbolisches EEG-Feature-Engineering für die Demenzklassifikation, und hängt die Leistung vom Validierungsprotokoll ab?

Methode

Proof-of-Concept-Studie mit öffentlich verfügbarem EEG-Demenzdatensatz; segmentweise 10-fold-CV, LOSO-CV und nested participant-grouped 10-fold-CV.

Ergebnisse

Segmentweise 10-fold-CV: 99,12–99,74 % Genauigkeit. LOSO: 51,19–63,81 %; nested participant-grouped: 49,00–62,50 %. Auf Teilnehmerebene lag LOSO bei 67,05 % für Demenz vs. gesund und bei 57,63 % vs. 61,02 % für AD vs. FD.

Einschränkungen

Segmentweise Validierung ließ Segmente derselben Person in Trainings- und Testfalten zu. Reproduzierbare krankheitsspezifische EEG-Merkmale und klinischer Nutzen wurden nicht belegt.

Schlussfolgerung der Autor:innen

Die Ergebnisse sprechen für die Machbarkeit interpretierbarer, algorithmisch abgeleiteter symbolischer Merkmalsrepräsentationen. Klinische Anwendung erfordert unabhängige Kohorten und bessere teilnehmerunabhängige Leistung.

Kernaussagen

  • Sehr hohe segmentweise Genauigkeit, aber starke Einbußen bei teilnehmerunabhängiger Validierung
  • DLob erzeugte Feature-Level-Visualisierungen der Kanalübergänge
  • Keine Belege für reproduzierbare krankheitsbezogene EEG-Merkmale oder klinische Diagnostik

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Clinical Medicine, online erschienen am 03.11.2026. Verfasst von Ferhat Balgetir und 10 weiteren Autor:innen. Beteiligte Einrichtung: Department of Neurology, School of Medicine, Firat University, 23119 Elazig, Türkiye. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objective

Dementia is a prevalent neurological disorder, and accurate detection is essential for timely intervention. This study aims to design an interpretable symbolic feature engineering framework for EEG-based dementia classification and to examine how its performance depends on the validation protocol.

Methods

We used a publicly available EEG dementia dataset including Alzheimer’s disease (AD), frontotemporal dementia (FD), and healthy controls. A self-organized explainable feature engineering (SOXFE) model was introduced with five phases: (i) feature extraction using Bipartite Channel Transformer (BCT) and Previous-Next Pattern (PNPat), (ii) feature selection via cumulative weight-based iterative neighborhood component analysis (CWINCA), (iii) explainable output generation using Directed Lobish (DLob), (iv) classification through an ensemble tkNN classifier, and (v) information fusion with combinational majority voting (CMV). Performance was evaluated with segment-wise 10-fold cross-validation, leave-one-subject-out (LOSO) cross-validation, and nested participant-grouped 10-fold cross-validation.

Results

Segment-wise 10-fold cross-validation, which allowed segments of one participant in both training and test folds, yielded accuracies of 99.12–99.74%. Participant-independent accuracy was substantially lower: 51.19–63.81% for LOSO and 49.00–62.50% for nested participant-grouped validation at the segment level. Participant-level LOSO accuracy equaled the majority-class rate for dementia versus healthy controls (67.05%) and fell below it for AD versus FD (57.63% versus 61.02%). No exact McNemar comparison showed an accuracy advantage over the majority-class baseline. DLob outputs were produced as feature-level visualizations.

Conclusions

The findings support the feasibility of interpretable, algorithm-derived symbolic feature representations. They do not establish reproducible disease-related EEG characteristics or clinical diagnostic utility. DLob outputs describe the channel transitions used by the model. Independent cohorts and stronger participant-independent performance are required before clinical application.

Schlagwörter

Stichwörter der Autor:innen

Directed LobishEEG signal classificationPNPatcombinational majority votingexplainable feature engineering

Autor:innen und Institutionen

  • Department of Neurology, School of Medicine, Firat University, 23119 Elazig, Türkiye.
  • Faculty of Engineering and Information Technology, University of Technology Sydney, Sydney, NSW 2007, Australia.
  • Australian International Institute of Higher Education, Brisbane, QLD 4000, Australia.
  • Department of Computer Engineering, College of Engineering, Erzurum Technical University, 25050 Erzurum, Türkiye.
  • School of Mathematics, Physics and Computing, University of Southern Queensland, Toowoomba, QLD 4350, Australia.
  • School of Computing and Information Technology, University of Wollongong, Wollongong, NSW 2522, Australia.

ORCID-Profile: Ferhat Balgetir, Prabal Datta Barua, Jun Shen, Padmanesan Narasimhan, Sengul Dogan, Burak Tasci, Mehmet Baygin, Turker Tuncer, Makiko Kobayashi

Zitieren

APA

Balgetir, F., Barua, P. D., Shen, J., Narasimhan, P., Shivshankar, S., Dogan, S., Tasci, I., Tasci, B., Baygin, M., Tuncer, T., & Kobayashi, M. (2026). Self-Organized Explainable Symbolic EEG Feature Engineering Using Directed Lobish and Previous-Next Pattern: A Proof-of-Concept Study on Dementia Classification. Journal of Clinical Medicine, 15(21), 8572. https://doi.org/10.3390/jcm15218572

Vancouver

Balgetir F, Barua PD, Shen J, Narasimhan P, Shivshankar S, Dogan S, et al. Self-Organized Explainable Symbolic EEG Feature Engineering Using Directed Lobish and Previous-Next Pattern: A Proof-of-Concept Study on Dementia Classification. Journal of Clinical Medicine. 2026;15(21):8572. doi:10.3390/jcm15218572

BibTeX
@article{Balgetir2026,
  author = {Balgetir, Ferhat and Barua, Prabal Datta and Shen, Jun and Narasimhan, Padmanesan and Shivshankar, Smitha and Dogan, Sengul and Tasci, Irem and Tasci, Burak and Baygin, Mehmet and Tuncer, Turker and Kobayashi, Makiko},
  title = {Self-Organized Explainable Symbolic EEG Feature Engineering Using Directed Lobish and Previous-Next Pattern: A Proof-of-Concept Study on Dementia Classification},
  journal = {Journal of Clinical Medicine},
  year = {2026},
  volume = {15},
  number = {21},
  pages = {8572},
  doi = {10.3390/jcm15218572}
}
Kostenlos starten →