PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
November 7, 2023Frontiers in Aging NeuroscienceOpen Access

Diagnosis of Alzheimer’s disease via resting-state EEG: integration of spectrum, complexity, and synchronization signal features

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Does the integration of spectrum, complexity, and synchronization EEG features accurately classify Alzheimer's disease versus normal subjects?

Population

36 AD subjects and 29 normal subjects from a public database

Comparison

Decision trees vs random forests vs SVM for categorizing resting-state EEG features in AD vs normal cases

Design

Machine learning classification study

Key result

The integration of spectrum, complexity, and synchronization features from resting-state EEG achieved a classification accuracy of 95.86% between Alzheimer's disease patients and healthy controls using a random forest algorithm.

Authors

XZXiaowei ZhengBWBozhi WangHLHao Liu

Discussion

Loading...

Member takes

Overview

May support EEG-ML screening research in AD; leaves open prospective validation before clinical adoption.

Study Design

Type

Observational (n=65)

Structured PICO

Does the integration of spectrum, complexity, and synchronization EEG features accurately classify Alzheimer's disease versus normal subjects?

P
Population
65 older adults (36 with Alzheimer's disease and 29 healthy controls) whose resting-state EEG recordings were analyzed to evaluate diagnostic classification.
E
Exposure
Integration of spectrum, complexity, and synchronization signal features of resting-state EEG using machine learning algorithms (decision trees, random forests, and support vector machine)
C
Comparator
Normal subjects (for classification) and comparison among different machine learning algorithms
O
Outcome
Classification accuracy between Alzheimer's disease and normal casessurrogate

Integrating spectrum, complexity, and synchronization features from resting-state EEG with machine learning algorithms yields high diagnostic accuracy for Alzheimer's disease.

Limitations

  • Only focused on the classification of AD and CN subjects, not the severity of AD.
  • Features were obtained by averaging EEG signals across all recorded electrodes, which may not capture specific brain region effects.
  • Regional distribution of the brain features corresponding to AD was not always consistent.

Cite This Study

Zheng et al. (2023) conducted an observational in Alzheimer's disease (n=65). Integration of spectrum, complexity, and synchronization EEG features vs. Healthy controls was evaluated on Classification accuracy between AD and normal cases. The integration of spectrum, complexity, and synchronization features from resting-state EEG achieved a classification accuracy of 95.86% between Alzheimer's disease patients and healthy controls using a random forest algorithm.

synapsesocial.com/papers/6aa2fb4a375ea2a2e2fddc3fhttps://doi.org/10.3389/fnagi.2023.1288295
View Full Paper
Ask AI
Bookmark
Share