Conception and evaluation of a musculoskeletal finite element model of the foot

Objective This work aims at developing a patient-specific Finite Element (FE) model of the foot in the context of predictive medicine. We propose here to include the realism of the anatomy by introducing all the detailed bony structures, ligaments and muscles segmented from the CT and MRI exams of the same patient. Material/patients and methods The model includes 30 bones modelled as articulated rigid-bodies connected with cables that simulate the 210 segmented ligaments in their actual positions and therefore define the articulations. The aponeurosis is modelled with fives parallel multipoint ligaments. A set of 15 extrinsic and intrinsic muscles based on linear axial muscle activation law is anatomically positioned in order to allow a natural movement of the foot. A FE mesh of the soft tissue was created by applying an automatic FE mesh generator to the surfaces resulting from MRI segmentation. The FE mesh has 95,883 elements and 56,806 nodes. Three soft tissue layers with Neo Hookean materials (Young moduli, Poisson Ratio) were created to represent a 1 mm skin layer (200, 0.485), the fat (30, 0.49) and muscle (60, 0.495) tissues. Results The foot model was evaluated by comparing its simulations with the pressure mat collected with the patient standing onto a Zebris FDM platform. For this, the patient weight was applied onto the foot model put in contact with a horizontal plate. The simulated pressures are close to the real ones (less than 2 N/cm 2 ) considering pressure measurement and subject weight uncertainties. Another evaluation of the model aimed at comparing the kinematics of the model resulting from simulated muscle activations with an actual 3D motion analysis of the patient foot coupled with an EMG monitoring of the tibialis posterior, tibialis anterior, soleus, medial gastrocnemius, lateral gastrocnemius and peroneus muscles. For an abduction/adduction 3D range of motion, the simulation is less than 4% different. Discussion – conclusion Globally, this foot model provides realistic simulations in both static and dynamic frameworks. In addition to its use for predictive medicine, it could become relevant for the simulation of neuro-orthopedic surgical interventions, pressure ulcer prevention, orthotic devices analysis or educational purposes.