MR physicist, anyone interested in using high permittivity materials to improve RF coil performanceThe introduction of relative High-Dielectric Constant (HDC) materials in radiofrequency (RF) coils has been shown to address efficiently the B₁⁺ inhomogeneity in ultra-high field MRI (B0≥3T) with limited Specific Absorption Ratio (SAR) constraints [1]. HDC materials act as a secondary RF field source modifying the global RF distribution in the transmit coil [2, 3]. Literature reports that they can be made out CaTiO₃ [1] or BaTiO₃ [4] powders mixed with de-ionized or deuterated water, or monolithic blocks of lead zirconium titanate (PZT) [5]. Unfortunately, none of them are really fulfilling the requirements to be used in high field clinical routine. HDC pads represent an important bulk in the coil, they reduce patient’s comfort, and they are aging rapidly. Some materials can be expensive and, like BaTiO₃, referenced as highly toxic [6]. To foster the RF transmission and avoid these inconveniences, we came up with an innovative solution based on the use of a new Meta-Atom (MA) structure, acting like a magnetic resonator in the RF coil, which solves all these limitations. To the knowledge of the authors, it is the first time that a MA structure is proposed for “dielectric shimming” purpose.The MA structure (pending patent) was designed in order to mimic the effect of a BaTiO₃ pad, currently one of the most efficient solution found in the literature [4, 7]. It was made of four 1mm diameter copper rods with a length of 40cm, separated from each other by 1,5cm, forming a rectangular parallelepiped in order to create magnetic mode [8,9]. The estimation of the equivalent dielectric constant of the MA structure, considered as a homogeneous object including also 2cm of vacuum representing the distance from the patient, has been estimated using CST Software (Darmstadt, Germany) and has been compared to a BaTiO₃ pad placed against the left ear measuring 12x10x1cm³ with a dielectric constant ε’=225. The pad permittivity has been determined using Nicolson and Ross protocol [10] associated to a de-embedding operation using an innovating coaxial cell with sample holder [11]. To assess the impact of our MA structure on the B₁⁺ distribution, validation experiments were performed using a birdcage head coil 1Tx/1Rx (Invivo Corp., Gainesville, USA) and a Specific Anthropomorphic Mannequin phantom (SPEAG, Zürich, Switzerland) on a 7T Magnetom MRI (Siemens Healthcare, Erlangen, Germany). B₁⁺ maps were acquired with an AFI sequence [12] with the MA structure or with a BaTiO₃ pad.Figure 1 shows the dielectric or equivalent dielectric constants of the different solutions. Their respective effects on the B₁⁺ distribution can be observed on Figure 2. Even further away from the object, the MA structure demonstrates a much stronger impact on the B₁⁺ field, spread over a larger volume. This is consistent with its design and its larger equivalent dielectric constant at 297MHz.Besides its clear effect on RF field distribution, our MA structure has the potential to overcome limitations of the dielectric pads currently used: it has a smaller size, and can be set further away from the patient. It is not aging, it is cost effective and very easy to manufacture. Its equivalent dielectric constant is also tunable modifying the length and the distance between the rods. Copper rods embedded in the MA structure are so far very long. A solution to shorten them is currently being developed. If so, it should be possible to integrate MA structures directly inside the transmission birdcage design, clearing space inside the coil. Thus, a combination of several MA structures inside the coil might not only restore locally the B₁⁺ excitation but also improve the global B₁⁺ NRMSE (Normalized Root Mean Square Error) in the whole brain at 7T. But before implementing such a solution in a state of the art routine clinical coil, some questions need to be addressed and will be the topic of later work. First, the interference of this MA structure with a specific receiving phased array must be examined. Furthermore, its impact on the SAR value needs to be evaluated accurately, even though the design of the structure has been elaborated in order to center the magnetic mode at 297MHz, reducing significantly the contribution of the electric mode.