IEEE Access (Jan 2024)
Microwave Bone Imaging: Reconstruction of Anthropomorphic Numerical Calcaneus Phantoms for Bone Diseases Diagnosis
Abstract
Microwave Tomography (MWT) is emerging as a potential imaging technique for monitoring bone health due to its ability to determine the dielectric properties of tissues under investigation. Studies have demonstrated a significant dielectric contrast between healthy and diseased human trabecular bones, which suggests that MWT has the potential to detect changes associated with bone disorders. The goal of this study was to assess the numerical reconstruction of an anthropomorphic two-layer (cortical and trabecular bone) calcaneus-shaped phantom, considering both healthy and diseased scenarios. Data were obtained using two different antenna configurations: a 16-antenna circular array and a 24-antenna calcaneus-shaped array. The electromagnetic (EM) inverse problem was solved by employing the Distorted Born Iterative Method (DBIM) in conjunction with the Iterative Method with Adaptive Thresholding for Compressed Sensing (IMATCS). An $L_{2}$ -based regularization method was applied along with the IMATCS. The results demonstrated that accurate reconstruction of various phantom properties at 1 GHz is feasible even under low signal-to-noise ratio (SNR) conditions. The achieved results suggest that the present approach can provide a good numerical reconstruction of the proposed phantoms both in the circular array and the calcaneus-shaped array scenarios, proving its robustness in distinguishing the overall reference and reconstructed dielectric properties with an average percentage difference of 10.43%. This work revealed the possibility of applying the suggested approach in a new scenario with a realistic phantom in two different antenna-array geometric configurations. In the future, this could be tested in an experimental context to assess the medical conditions of the human bones.
Keywords
