Best methods and data to reconstruct paediatric lower limb bones for musculoskeletal modelling

Area of Science:

• Biomechanical Engineering
• Medical Imaging
• Pediatric Orthopedics

Background:

• Generic musculoskeletal models often oversimplify pediatric anatomy, potentially affecting biomechanical simulation accuracy.
• Subject-specific models offer realistic skeletal representations but require efficient reconstruction methods.
• Previous studies successfully applied statistical shape modeling (SSM) and morphing to adult bones but not pediatric data.

Purpose of the Study:

• To reconstruct pediatric pelvis, femur, and tibiofibular bones using the Musculoskeletal Atlas Project (MAP) Client.
• To evaluate the accuracy of 10 different SSM and morphing pipelines for pediatric bone reconstruction.
• To compare the accuracy of image-based reconstructions with generic, linearly scaled models.

Main Methods:

• Reconstruction of pediatric bones (pelvis, femurs, tibiofibular) from 18 children using the MAP Client with MRI and/or motion capture data.
• Development and testing of 10 SSM and morphing pipelines.
• Quantification of reconstruction accuracy using Jaccard index, root mean square distance error, and Hausdorff distance.
• Comparison with linearly scaled generic bone geometries from the OpenSim gait2392 model.

Main Results:

• Pipelines utilizing MRI data (alone or with motion capture) achieved accurate pediatric bone reconstructions (Jaccard index > 0.8).
• Linearly scaled OpenSim geometries yielded the least accurate reconstructions.
• Principal component-based scaling methods demonstrated size-dependent accuracy, performing poorly for smaller children.

Conclusions:

• Image-based SSM and morphing techniques, particularly those informed by MRI data, are effective for accurate pediatric lower limb bone reconstruction.
• Generic, scaled models are less suitable for pediatric biomechanical simulations due to inaccuracies.
• Accurate subject-specific pediatric musculoskeletal models are crucial for reliable biomechanical analysis in children.