Determining bone position from wearable antennas using microwave imaging: A feasibility study

Imaging of the bone is an important clinical tool in detecting fractures, determining loss of bone density and in improving movement and gait analysis, with visualisation predominantly performed using ionising imaging. Microwave imaging, an alternative non-ionising imaging modality, has shown promising results in analysing bone density variation and in determining the presence of tears in joint tissues. In this study, we aimed to detect the location of the bone in the leg using radar-based microwave imaging. Confocal imaging algorithms were applied to scattering data acquired from simulated wearable antennas on a male human model without the use of a coupling liquid. We successfully detected both thigh and shin bones, with a localisation error of 2.5 cm, 0.91 cm and 1.34 cm for the femur, tibia and fibula centres respectively. Notably, these errors are in line with tumour detection errors. Our methodology and results illustrate a safe, easy-to-implement and simple pipeline using off-the-shelf antennas and algorithms to determine bone position using wearable sensors. We believe this technique has wide ranging applications, particularly in improving the accuracy of clinical movement analysis systems.