Synthetic data analysis for early detection of Alzheimer progression through machine learning algorithms

Area of Science:

• Neuroscience
• Artificial Intelligence
• Medical Diagnostics

Background:

• Alzheimer's disease (AD) is a progressive neurodegenerative disorder causing irreversible neuronal damage.
• Early detection is crucial for managing AD symptoms and slowing progression.
• Current diagnostic methods can be improved for earlier and more accurate identification.

Purpose of the Study:

• To propose an early detection model for Alzheimer's disease (AD) using machine learning.
• To analyze Alzheimer's progression patient datasets from the Alzheimer's Disease Neuroimaging Initiative (ADNI).
• To build an ensemble model for early diagnosis based on neuropsychological assessments.

Main Methods:

• Utilized neuropsychological assessment data from ADNI.
• Employed feature selection techniques: Recursive Feature Elimination (RFE) and Akaike Information Criterion (AIC).
• Developed an ensemble machine learning model combining Logistic Regression (LR), Artificial Neural Networks (ANN), Support Vector Machines (SVM), K-Nearest Neighbors (KNN), and Nearest Centroid (Nearcent).

Main Results:

• The study analyzed 13 distinct patient subsets, comparing various cognitive states (CN, SMC, EMCI, LMCI, AD).
• A customized model was created using RFE and AIC for feature selection and model optimization.
• The ensemble model demonstrated potential for early AD detection based on neuropsychological data.

Conclusions:

• The developed ensemble model offers a novel approach for early Alzheimer's disease diagnosis.
• This model can be integrated into a backend platform for one-versus-all analysis.
• The findings provide a basis for earlier diagnosis and intervention in Alzheimer's disease.