The design of a MRI-compatible knee loading device

Abstract

The knee is the most common joint affected by osteoarthritis (OA). Evaluating knee cartilage in MRI is usually done under non-loading conditions. However, load-bearing is a key capacity of cartilage, so using compression may be a tool for clinicians to identify early OA. A loading device could simulate weight-bearing conditions in a supine position. This report presents the development of an MRI-compatible knee loading device, following the double diamond method and the V-model.

The most important requirement are that the device must apply an axial, accurate, and adjustable load through the foot to the patient’s knee in a supine position. The device should be able to apply different small knee flexion angles. Existing loading devices, MRI procedures, knee biomechanics, and potential risks were analyzed to establish a comprehensive set of requirements. A knee loading device has been developed where a balance needed to be found between feasibility, functionality and complexity, with feasibility as the main priority to ensure a working foundation for further enhancements.

A biomechanical analysis was conducted to understand the impact of applying compression to the knee in a supine position on the internal joint moments, which reflects the effort that needs to be provided by the patient. Adjustments in knee flexion angle and the height of the applied load at the foot were made to evaluate their effect. The results showed that a 10° knee flexion with a load application through the ankle requires the least effort from the patient, making it the most feasible loading condition to maintain stable for several minutes. For the design, this meant that the height of the footplate was set twice the distance between the bottom of the calcaneus and the position where the load goes through the ankle. Although 10° knee flexion might the most feasible loading condition, it does not necessarily mean it is
the best indicator for early OA. Therefore, the device still needs to enable different knee flexion angles.

The load is applied at the foot through a footplate, which slides along axes. To generate the load, elongating elastic material was chosen because of its practicality and safety. Suspended weights were excluded, as they require an inconvenient set-up and are not safe in case of emergency. Since a setup with suspended weights requires more space and covers both the patient bed and MRI table, it is not possible to disconnect the MRI table from the MRI scanner rapidly. The final design of the MRI knee loader ensures the patient can be positioned, and the load could be applied without having to change the position. To make the load generation feasible for radiology laboratory personnel, a driving wheel and pulley systems were incorporated to reduce the force required at the wheel. Additionally, a knee support is implemented to apply a knee flexion angle to the patient.

A prototype has been manufactured to evaluate the functional application of the applied load and knee flexion angle, as well as patient compatibility. It included the essential functional components: the footplate, knee support and the plates they are connected to. The prototype demonstrated the capability of applying a load to the foot in an experimental set-up. A linear relation was found between the input force and the measured force at the footplate, mimicking the load that is applied at the foot. Unfortunately, the measured force was not sufficiently accurate and the variation between the repeated measurements was found substantial. It was argued that this was due to the friction in the current design.

Further design improvements are required to overcome the limitations of this design. Compromises were done to improve feasibility, but it could be analysed how the functionality and accuracy could be improved. A starting point could be to find a better fit between the bearings in the footplate and the axes, to reduce friction and ensure a more accurate application of the load. The footplate should be redesigned so it does not clamp. Moreover, the complete final design of the MRI knee loader (including the footplate, knee support and their connection plate, as manufactured in the prototype, as well as the driving wheel and the table with the double bottom embedding the cable and elastic) must be produced to verify and validate the device’s overall functionality, patient compatibility, suitability within the MRI
environment and workflows, and operational ability by the MRI laboratory personnel.