Quantifying Patient-Specific Plantar Soft Tissue Stiffnessusing Magnetic Resonance Imaging and a HydraulicLoading Device

Abstract

Diabetes mellitus is a prevalent health concern in the United States that has many adverse symptoms, one of which is increased plantar soft tissue stiffness. This change in the mechanical properties of foot tissue has been hypothesized to lead to large internal peak stresses that could cause ulcers. Ulcers can lead to foot tissue infections and eventually amputations. Of all non-traumatic lower limb amputations for subjects greater than 20 years old, 60% follow this pathology. It is important to increase our understanding of the stress distribution in the foot under various conditions, such as loading rate/magnitude and tissue healthiness, in order to improve computational foot models and predictions for in vivo subjects with and without diabetes. This study aims to achieve this by combining three-dimensional (3D) internal tissue deformation imaging (gated MRI) with a non-metallic hydraulic loading device (the hydraulic plantar soft tissue reducer – HyPSTR). There were two major aspects of the HyPSTR project. The first was part to design, build, and verify an MRI-compatible hydraulic device to load a subject’s foot in a controlled manner within the MRI’s magnetic core. Preliminary testing was performed with load cells and rigid aluminum jigs to ensure the motor and piston design was capable of producing physiologic loading conditions. The second aspect was to use the HyPSTR with ultrasound (pilot testing) and gated MRI protocols to capture deformations under dynamic forces. The data collected were processed to generate stiffness curves and subsequently used with inverse finite element analysis (FEA). The inverse FEA modeling was done by Vara Isvilanonda for his PhD dissertation in 2015. The construction and verification of the HyPSTR was a significant undertaking and highlighted several issues with the hydraulic performance and jig used to secure the subjects’ legs when under load. Primarily, the loading platen was only able to operate at a maximum of 0.2 Hz, which is slower than the frequency content of gait (the goal rates were 1 Hz to 6 Hz). This problem was caused by a water hammer (pressure ringing) effect due to the rapid momentum changes when the hydraulics reversed direction (load vs unload movement). Additionally, the load magnitudes were only approximately 25% of subject body weight. Discomfort on the dorsal aspect of the foot, at the interface with an ankle foot orthosis (AFO), prevented testing with larger forces. Even with these problems, the study tested one subject with diabetes and one without, and demonstrated that diabetic plantar soft tissue was stiffer than healthy tissue. These results are only for two subjects, but they provide cursory proof that the HyPSTR can measure patient-specific stiffnesses. Furthermore, the procedures and devices include all of the functionality needed to make in vivo, patient-specific, dynamic, 3D deformation measurements. The next challenge is not to add more components, but rather to improve the control and range of what already exists.