Frontiers in Medicine, Год журнала: 2025, Номер 12
Опубликована: Фев. 19, 2025
The prevalence of Alzheimer's disease (AD) poses a significant public health challenge. Distinguishing AD stages remains complex process due to ambiguous variability within and across stages. Manual classification such multifaceted massive data brain volumes is operationally inefficient vulnerable human errors. Here, we propose precise systematic framework for classification. core this discovers analyzes functional connectivity among regions interest (ROIs) brain. Multivariate Pattern Analysis (MVPA) applied extract features that reveal patterns in the These are then used as inputs an Extreme Learning Machine (ELM) model classify model's performance assessed through comprehensive evaluation metrics ensure robustness reliability. Applying on datasets which contain meticulously validated fMRI scans OASIS Neuroimaging Initiative datasets, validate merit proposed work. framework's results show improvement collective two-class multi-class Feeding ELM with MVPA yield decent outcomes given generalizable computationally-efficient model. This study underscores effectiveness approach accurately distinguishing stages, offering potential improvements detection.
Язык: Английский