Antoine Motte dit Falisse

SimCP: A Simulation Platform to Predict Gait Performance Following Orthopedic Intervention in Children with Cerebral Palsy

We present a simulation platform that will enable clinicians to evaluate the effect of different treatment options on gait performance in children with cerebral palsy (CP) in order to select the treatment with the highest potential to normalize the patient’s gait pattern. We present a case study to demonstrate the use of the platform. We created a neuro-musculoskeletal model of a 10-year old female child with mild spastic triplegic CP (GMFCS II) who was treated with single-event multilevel surgery based on medical imaging and motion capture data collected before the surgery. Based on this model, we predicted that the treatment would reduce the capability gap, i.e. the torque deficit of the patient with respect to the joint torques needed for normal walking. This prediction was in accordance with the closer-to-normal post-treatment gait kinetics of the child.

The influence of maximum isometric muscle force scaling on estimated muscle forces from musculoskeletal models of children with cerebral palsy

BACKGROUND Musculoskeletal models do not include patient-specific muscle forces but rely on a scaled generic model, with muscle forces left unscaled in most cases. However, to use musculoskeletal simulations to inform clinical decision-making in children with cerebral palsy (CP), inclusion of subject-specific muscle forces is of utmost importance in order to represent each childs compensation mechanisms introduced through muscle weakness. RESEARCH AIM The aims of this study were to (i) evaluate if maximum isometric muscle forces (MIMF) in musculoskeletal models of children with CP can be scaled based on strength measurements obtained with a hand-held-dynamometer (HHD), (ii) evaluate the impact of the HHD based scaling approach and previously published MIMF scaling methods on computed muscle forces during gait, and (iii) compare maximum muscle forces during gait between CP and typically developing (TD) children. METHODS Strength and motion capture data of six CP and motion capture data of six TD children were collected. The HHD measurements to obtain hip, knee and ankle muscle strength were simulated in OpenSim and used to modify MIMF of the 2392-OpenSim model. These muscle forces were compared to the MIMF scaled on the childs body mass and a scaling approach, which included the body mass and muscle-tendon lengths. OpenSim was used to calculate peak muscle forces during gait. RESULTS Ankle muscle strength was insufficient to reproduce joint moments during walking when MIMF were scaled based on HHD. During gait, peak hip and knee extensor muscle forces were higher and peak ankle dorsi-flexor forces were lower in CP compared to TD participants. SIGNIFICANCE HHD measurements can be used to scale MIMF for the hip and knee muscle groups but underestimate the force capacity of the ankle muscle groups during walking. Muscle-tendon-length and mass based scaling methods affected muscle activations but had little influence on peak muscle forces during gait.BACKGROUND Musculoskeletal models do not include patient-specific muscle forces but rely on a scaled generic model, with muscle forces left unscaled in most cases. However, to use musculoskeletal simulations to inform clinical decision-making in children with cerebral palsy (CP), inclusion of subject-specific muscle forces is of utmost importance in order to represent each childs compensation mechanisms introduced through muscle weakness. RESEARCH AIM The aims of this study were to (i) evaluate if maximum isometric muscle forces (MIMF) in musculoskeletal models of children with CP can be scaled based on strength measurements obtained with a hand-held-dynamometer (HHD), (ii) evaluate the impact of the HHD based scaling approach and previously published MIMF scaling methods on computed muscle forces during gait, and (iii) compare maximum muscle forces during gait between CP and typically developing (TD) children. METHODS Strength and motion capture data of six CP and motion capture data of six TD children were collected. The HHD measurements to obtain hip, knee and ankle muscle strength were simulated in OpenSim and used to modify MIMF of the 2392-OpenSim model. These muscle forces were compared to the MIMF scaled on the childs body mass and a scaling approach, which included the body mass and muscle-tendon lengths. OpenSim was used to calculate peak muscle forces during gait. RESULTS Ankle muscle strength was insufficient to reproduce joint moments during walking when MIMF were scaled based on HHD. During gait, peak hip and knee extensor muscle forces were higher and peak ankle dorsi-flexor forces were lower in CP compared to TD participants. SIGNIFICANCE HHD measurements can be used to scale MIMF for the hip and knee muscle groups but underestimate the force capacity of the ankle muscle groups during walking. Muscle-tendon-length and mass based scaling methods affected muscle activations but had little influence on peak muscle forces during gait.

O 107 – Impact of subject-specific musculoskeletal geometry on estimated joint kinematics, joint kinetics and muscle forces in typically developing children

1.. Introduction: Gait analysis together with musculoskeletal modeling can be used to calculate muscle forces and assess pathological gait [1]. No generic, pediatric musculoskeletal models are available and, therefore, linear scaling methods are commonly used to personalize a generic, adult musculoskeletal model to the child’s anthropometry. 2. Research: How different are joint kinematics, joint kinetics and muscle force estimates of generic scaled models compared to medical-imaging based models in typically developing (TD) children? 3. Methods: 3D motion capture data and magnetic resonance images (MRI) of a TD boy (age: 8 years; height: 1.23 m; weight: 20.4 kg) were collected. Two musculoskeletal OpenSim models were created: (1) a scaled generic model (M_gen), and (2) a MRI-based model, which included subject-specific musculoskeletal geometry (M_mri) [2]. Joint kinematics, joint kinetics and muscle forces were calculated for each model using OpenSim 3.3 [3]. Joint kinematics, joint kinetics, muscle force waveforms, as well as femoral anteversion angle, neck-shaft angle and hip joint centre location were compared between both models. 4. Results: Joint kinematics and joint kinetics were surprisingly similar between the M_gen and M_mri with root-mean-square-differences below 2.8° and 0.05Nm/kg for joint angles and moments, respectively (Fig. 1, Fig. 2). Depending on the analyzed muscle, differences in muscle forces varied substantially (up to 230% difference) between the M_gen and M_mri (Fig. 3). Femoral anteversion and neck-shaft angles differed between M_gen and M_mri by 12 and 5 degrees, respectively. The hip joint centre position differed between both models by 5, 15 and 6 mm in the anterior/posterior, superior/inferior and medial/lateral direction, respectively.