Related Experiment Video
Updated: Apr 26, 2026
07:30
3D Printing Model of a Patient's Specific Lumbar Vertebra
Published on: April 14, 2023
2.7K
A data driven model for optimal orthosis selection in children with cerebral palsy
Andrew J Ries1, Tom F Novacheck2, Michael H Schwartz2
1Gillette Children's Specialty Healthcare, St. Paul, MN, United States; Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN, United States.
Gait & Posture
|July 29, 2014
Summary
A new Random Forest Algorithm (RFA) model accurately predicts optimal orthosis selection for individuals with diplegic cerebral palsy (CP). This AI-driven approach significantly improves gait outcomes compared to current methods.
Area of Science:
- Biomedical Engineering
- Rehabilitation Technology
- Clinical Biomechanics
Background:
- Diplegic cerebral palsy (CP) often requires orthotic intervention to improve gait.
- Current orthosis selection relies on clinical expertise, with variable outcomes.
- Optimizing orthosis choice is crucial for enhancing mobility and function in CP patients.
Purpose of the Study:
- To develop and evaluate a statistical model for selecting the optimal orthosis for individuals with diplegic CP.
- To assess the potential clinical benefit of an AI-driven orthosis selection model.
- To compare the model's predictions against existing orthosis prescriptions.
Main Methods:
- A Random Forest Algorithm (RFA) model was trained on retrospective data from 476 individuals with diplegic CP.
- The model predicted optimal orthosis type (SAFO, PLS, HAFO, SMO, FO) based on gait outcome (ΔGDI).
- Model performance and clinical benefit were evaluated in a cohort of 1016 individuals.
Main Results:
- The RFA model showed limited agreement (14%) with existing orthosis prescriptions.
- For 56% of limbs, the model correctly predicted no orthotic benefit.
- Predicted average ΔGDI improved from +0.4 (current care) to +5.6 using the model.
Conclusions:
- An RFA-based orthosis selection model can significantly enhance gait outcomes for diplegic CP.
- The model offers a data-driven approach to personalized orthotic management.
- Further validation through multi-center and prospective studies is recommended.
