A comprehensive biomechanical phenotyping framework for diabetic foot ulcer risk stratification using multi-modal...

Area of Science:

• Biomechanics
• Machine Learning
• Diabetic Foot Care

Background:

• Current diabetic foot ulcer risk assessment lacks precision in identifying high-risk biomechanical phenotypes.
• There is a need for enhanced ulcer risk stratification methods.

Purpose of the Study:

• To develop a comprehensive biomechanical profiling framework integrating multi-modal gait analysis with machine learning.
• To enhance diabetic foot ulcer risk stratification.

Main Methods:

• Prospective cross-sectional study of 214 participants (diabetic foot ulcer patients, diabetic controls, healthy controls).
• Multi-modal gait analysis including plantar pressure mapping, wearable inertial sensors, 3D motion capture, and electromyography.
• Machine learning (unsupervised for phenotyping, supervised for prediction) with nested cross-validation.

Main Results:

• Diabetic foot ulcer patients showed elevated forefoot pressures and cautious gait patterns.
• Four distinct biomechanical phenotypes with differential ulceration risks (OR: 3.2-8.7) were identified.
• A random forest model achieved 94.3% accuracy in classifying diabetic foot ulcer risk using six biomechanical features, outperforming conventional methods. Unrecognized instability patterns predicted ulcer development 6-8 months prior.

Conclusions:

• Novel biomechanical phenotypes with differential ulcer susceptibility were identified and validated.
• Integration of machine learning and multi-modal gait analysis enables precise risk stratification.
• This represents a paradigm shift towards proactive, phenotype-specific diabetic foot care.