Introduction:Early identification of tumor-like changes in bone remains a major diagnostic challenge because conventional imaging methods mainly rely on structural contrast and may involve ionizing radiation, high cost, or limited point-of-care accessibility. In line with the growing interest in bioelectronic and nanoelectronic biosensor technologies for translational diagnostics, this study investigates a low-cost, non-ionizing microwave sensing platform for detecting dielectric changes associated with a tumor-mimicking bone phantom. Methods:A compact microstrip resonator antenna was designed in CST Microwave Suite and fabricated on an FR4 substrate for localized near-field sensing. A multilayer bone phantom containing cortical bone- and marrow-mimicking regions was prepared using wheat flour, deionized water, dextrose, and olive oil. A separate water-rich gelatin-based tumor phantom was prepared to reproduce the higher dielectric response expected from malignant tissue. The dielectric properties of the bone and tumor phantoms were measured using a Vector Network Analyzer (VNA)-based open-ended coaxial probe system. The fabricated antenna was then experimentally evaluated by reflection coefficient (S11) measurements at two healthy phantom positions and one tumor-over-phantom position. Results:Dielectric characterization confirmed a clear contrast between the bone phantom and tumor-like region in both the real and imaginary parts of relative permittivity. The tumor-loaded configuration produced a downward resonance shift of 110 MHz around 4.2-4.3 GHz, relative to the healthy reference, which was 2.75 times larger than the variation between the two healthy positions. In addition, a 4.47 dB change in S11 magnitude was observed, confirming that the tumor-like inclusion altered the near-field dielectric loading and impedance-matching condition of the resonator. Conclusions:The proposed microstrip resonator antenna demonstrates the feasibility of a compact RF bioelectronic sensing approach for detecting tumor-like dielectric perturbations in a controlled bone phantom environment. Although not intended as a clinical diagnostic device at this stage, the results support further development of this platform toward translational microwave biosensing, more realistic preclinical phantoms, array-based configurations, and AI-assisted classification for future diagnostic applications.
