Clinical Gait Analysis: Characterizing Normal Gait and Pathological Deviations Due to Neurological Diseases

Area of Science:

• Biomechanics
• Neurology
• Data Science

Background:

• Characterizing normal gait and pathological deviations is crucial for diagnosing neurological disorders.
• Current methods often rely on average gait patterns, potentially missing individual variations.

Purpose of the Study:

• To develop an unsupervised approach for characterizing normal gait profiles (NGPs) using real gait cycles.
• To quantify pathological gait deviations from these NGPs.
• To stratify deviations and assess motor impairment in neurological conditions.

Main Methods:

• Unsupervised learning using Dynamic Time Warping (DTW) to identify distinct normal gait profiles (NGPs).
• Utilizing NGPs to measure deviations in pathological gait cycles via DTW.
• Applying Hierarchical Clustering to categorize gait deviations.

Main Results:

• Three distinct NGPs are identified as necessary for accurately characterizing normal gait heterogeneity.
• The method effectively quantifies pathological gait deviations, identifying affected limbs in hemiplegic patients and assessing impairment severity in paraplegic and tetraplegic patients.
• Raw gait signals, without normalization, preserve crucial dynamic information for accurate deviation measurement.

Conclusions:

• The proposed methodology offers a robust way to analyze gait abnormalities in neurological diseases.
• It provides a sensitive tool for quantifying motor impairment and potentially evaluating rehabilitation therapies.
• Preserving temporal dynamics in gait signals is vital for accurate pathological assessment.