TToTP model: An automated detection of Alzheimer's disease and frontotemporal dementia using Tesla Tree of Thought Pattern with EEG signals
Background Alzheimer's disease (AD) and frontotemporal dementia (FTD) are common neurological disorders. Objective: To facilitate the early detection of these conditions, we aimed to propose a new self-organized explainable feature engineering (SOXFE) model. Additionally, we sought to investigate the feature extraction capabilities. Methods We used a publicly available electroencephalography (EEG) dementia detection dataset. This dataset was created to collect EEG signals from participants in three categories: (i) AD, (ii) FTD, and (iii) healthy control (HC). To automatically detect these groups, we proposed a new SOXFE model. This model includes a novel feature extraction function inspired by Tesla's 369 vortex and the Tree of Thought (ToT), termed the Tesla Tree of Thought Pattern (TToTP). In the feature selection phase, we applied cumulative weighted iterative neighborhood component analysis (CWINCA), to select the most informative features. In the classification phase, we employed the t algorithm-based k-nearest neighbors (tkNN). The fourth phase involves information fusion, where the results from each channel are voted on using iterative majority voting (IMV), and the best outcome is selected. The final phase is explainable results generation, in which a cortical channel map (CCMap) is created based on the findings from the information fusion phase. Results We applied the TToTP-based SOXFE model to the EEG dementia dataset, achieving over 99% classification accuracy. Additionally, a cortical channel map was generated using the classification results. Conclusions The proposed TToTP-based SOXFE model achieved high classification performance and successfully generated a cortical channel map. The model contributes to both feature engineering and neuroscience
Authors
- Abdul Hafeez‐Baig (ORCID: https://orcid.org/0000-0003-3848-8008)
- Prabal Datta Barua (ORCID: https://orcid.org/0000-0001-5117-8333)
- Ru‐San Tan (ORCID: https://orcid.org/0000-0003-2086-6517)
- Şengül Doğan (ORCID: https://orcid.org/0000-0001-9677-5684)
- İrem Taşçı (ORCID: https://orcid.org/0000-0001-7069-769X)
- Türker Tuncer (ORCID: https://orcid.org/0000-0002-5126-6445)
- Burak Taşçı (ORCID: https://orcid.org/0000-0002-4490-0946)
- U. R. Acharya
Institutions
- Fırat University (TR)
- University of Southern Queensland (AU)
- Duke-NUS Medical School (SG)
- National Heart Centre Singapore (SG)
Publication Details
- Journal
- Journal of Alzheimer s Disease
- Published
- 2026-11-21
- DOI
- https://doi.org/10.1177/13872877261509096
- Primary Topic
- EEG and Brain-Computer Interfaces
- Type
- article
- Field-Weighted Citation Impact
- 0.00