Gait Rhythm Dynamics for Neuro-Degenerative Disease Classification via Persistence Landscape- Based Topological...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Data Science

Background:

• Neuro-degenerative diseases are progressive nervous system disorders with significant clinical impact.
• Gait rhythm dynamics analysis is crucial for assessing disease progression and patient quality of life.
• Stride-to-stride fluctuations in gait dynamics can quantify pathological alterations in locomotor control.

Purpose of the Study:

• To develop and validate a topological data analysis-inspired nonlinear framework for gait dynamics analysis.
• To utilize persistence landscapes as input for classifiers to differentiate neuro-degenerative diseases from healthy controls.
• To compare gait dynamics in healthy controls (HC) with those in patients with amyotrophic lateral sclerosis (ALS), Huntington's disease (HD), and Parkinson's disease (PD).

Main Methods:

• Applied a nonlinear framework inspired by algebraic topology theory to analyze gait dynamics.
• Employed topological representations, specifically persistence landscapes, as features for machine learning classifiers.
• Compared stride-to-stride time series data from HC, ALS, HD, and PD patient groups.

Main Results:

• The proposed topological framework demonstrated high accuracy in discriminating between healthy subjects and patients with neuro-degenerative diseases.
• The methodology effectively visualized gait dynamics, enabling classification of different neuro-degenerative conditions.
• Results indicate the potential of gait rhythm analysis using topological features for disease identification.

Conclusions:

• This study presents the first topological representation-based method for classifying neuro-degenerative diseases using gait rhythm data.
• The approach offers a novel way to visualize and quantify gait dynamics for diagnostic purposes.
• The method holds potential for earlier disease intervention and continuous patient state monitoring.