AI-Based Severity Classification of Dementia Using Gait Analysis

Area of Science:

• Computational neuroscience
• Biomedical engineering
• Artificial intelligence in healthcare

Background:

• Dementia severity classification is challenging due to complex, high-dimensional gait data.
• Conventional statistical methods may fail to capture subtle patterns in gait analysis.
• Artificial intelligence (AI) and machine learning (ML) offer potential solutions for complex data analysis.

Purpose of the Study:

• To explore the utility of AI in classifying dementia severity via gait analysis.
• To examine how ML addresses limitations of traditional statistical approaches in this domain.
• To identify key gait features for dementia severity classification using ML techniques.

Main Methods:

• Utilized gait analysis data from 34 mild cognitive impairment (MCI), 25 mild dementia, 26 moderate dementia patients, and 54 healthy controls.
• Employed a support vector machine (SVM) classifier for dementia severity categorization.
• Applied machine learning techniques, including principal component analysis (PCA) and gradient-based feature selection, for dimensionality reduction and feature identification.

Main Results:

• Machine learning effectively identified key gait features relevant to dementia severity.
• ML techniques successfully handled high-dimensional gait data, complementing traditional statistical analyses.
• AI-based tools demonstrated promise in uncovering subtle gait patterns indicative of dementia severity.

Conclusions:

• AI and ML show significant utility in classifying dementia severity using gait analysis data.
• ML methods can overcome limitations of conventional statistics in analyzing complex, high-dimensional datasets.
• Findings support the development of advanced AI-driven diagnostic models for dementia.