Magnetic resonance imaging (MRI) is an enabling technology platform for imaging applications. In MRI, the imaging frequency falls within the radio frequency (RF) range where the tissue absorption of electromagnetic power is conveniently very low (e.g., compared to X-ray imaging), making MRI medically safe. As a result, MRI has evolved into a major imaging tool in medicine. However, in MRI, it is typically difficult to receive a magnetic resonance signal from tissue near a metallic implant, which hinders imaging of the implant device neighborhood to observe, monitor, and make assessment of the recovery and tissue compatibility. This can be accomplished by using locally resonating implants, but such implantable local resonators compatible with MRI that simultaneously feature reasonable chip size are currently not available (although there are some MRI-guided catheter applications). In this thesis, we proposed and developed a new class of implantable chip-scale local resonators that operate at radio frequencies of MRI, despite their small size, for the purposes of enhancing the signal-to-noise ratio (SNR) and thus the resolution in their vicinity. Here we addressed the scientific challenge of achieving low resonance frequency while maintaining chip-scale size suitable for potential MR-compatible implants. Using only biocompatible materials (gold, nitrides, and silicon or polyimide) within a substantially reduced footprint (miniaturized by 2 orders of magnitude), we demonstrated novel chip-scale designs based on the basic concept of split ring resonators (SRRs). Different than classical SRRs or those loaded with lumped elements (e.g., thin-film lumped loading), however, in our designs we loaded the SRR geometry in a distributive manner with a micro-fabricated dielectric thin-film layer to increase effective capacitance. For a proof-of-concept demonstration, we fabricated 20 mm ´ 20 mm resonators that operate at the resonance frequency of 130 MHz (compatible with 3 T MRI system) when distributively loaded with the capacitive film, which would otherwise operate around 1.2 GHz as a classical SRR of the same size if not loaded. It is worth noting that this resonance frequency of 130 MHz would normally require a classical SRR of 20 cm ´ 20 cm, a chip size 100-fold larger than ours. Designing and fabricating flexible thin-film resonators, we also showed that this architecture can be tuned by bending and is appropriate for non-planar surfaces, which is often the case for in vivo imaging. The phantom images indicated that, depending on the resonator configuration, these novel self-resonating structures increase SNR of the received signal by a maximum factor of 4 to 150 and over an enhancement penetration up to 10 mm into the phantom. This corresponds to a resolution enhancement in the 2D image by a factor of 2 to 12, respectively, under the same RF power. These in vitro experiments prove that it is possible to operate our local resonators at reduced frequencies via the help of distributive loading on the same chip. These findings suggest that proposed implantable resonator chips make promising candidates for self-resonating MR-compatible implants.