RF field measurement under MRI presents a significant measurement challenge. Both components of the RF field, the electric field and the magnetic field, are uniquely important for real-time sensing during an MR-sequence: the electric field contributes to deposited RF power in the tissue, leading to increased tissue temperature, while the magnetic field is used for imaging and localization. Conventional RF field sensors use long electrical conductors for signal transmission, which distort images and contribute to local RF power hot-spots, potentially damaging tissue on or in the patient. Current MRI-compatible sensors employ transduction from electrical to optical signals for RF immune signal transmission, which subverts these issues. However, demonstrated sensors either utilize complex, actively powered components, or rely on large, rigid components for passive sensing, all of which may constrain the device geometry, performance, and application. The feasibility of an acousto-optical (AO) RF field sensor structure based on light intensity modulation via an antenna-coupled piezoelectric transducer modulating a fiber-Bragg grating (FBG) was demonstrated earlier. This work develops and validates an end-to-end model of the AO sensor to enable design optimization, and a fabrication process based on conformal piezo-composite polymer transducers over optical fibers. The use of coaxial piezoelectric transducers increases the acoustic coupling efficiency and minimize the size of the sensor for seamless integration on interventional MRI catheters as compared to planar and bulky piezo crystals glued to optical fibers. Sensors are designed and fabricated for both E-field and B-field sensing capabilities under 3T MRI. These devices are also used to validate the complete electromechanical model for the antenna-coupled coaxial transducer to predict the sensor response to arbitrary incident RF fields. Experiments under MRI show that the sensitivity of fabricated devices is comparable to the previous FBG-based sensor and confirm the feasibility of fully RF-immune device tracking under MRI.