Introduction

Recent developments in additive manufacturing and data analytics have significantly advanced patient-specific medical devices in the healthcare sector. Additive manufacturing or 3D-printing techniques can be used to fabricate subject-specific MRI hardware components. Unlike conventional manufacturing techniques, additional geometric complexity in design does not translate into additional manufacturing costs or time. More importantly, the seamless transition between digital data and physical objects allows for integrating electromagnetic simulations prior to fabrication, facilitating design iterations.¹ Patient-induced B₀ inhomogeneities are an important factor significantly affecting the imaging quality in terms of signal-to-noise ratio (SNR) and geometric distortion.^2,3 Since anatomical features are highly subject-specific, so are the type and extent of the magnetic field distortions. Traditional passive shims are constructed by placing ferro-, dia- or paramagnetic materials at discrete locations inside the magnet bore in an iterative process.^2,4 This time-consuming process has difficulties rendering the higher-order spherical harmonics (SHs) terms required to shim field inhomogeneities induced by complex anatomical features. We developed a powder-binder jetting technique to manufacture subject-specific passive shims. The method deposits variable concentrations of magnetic ink at precise pre-calculated positions in a patient-specific shim design to sculpt B₀ towards a target distribution via the passive response of the 3DP shims to B₀. Shim designs producing discrete SH terms were fabricated as a proof-of-concept towards subject-specific passive shims. A poor geometrical fit between the RF coil and subject also limits SNR.⁵ Several approaches relying on flexible coil substrates have shown promising results.^6–9 However, the flexibility hinders the precise and reproducible coil placement on the subject. We developed three additive manufacturing techniques for the fabrication of patient-specific RF coils. Proof-of-concept 3D-printed subject-conform coils were fabricated to image a human thumb's metacarpophalangeal (MCP) joint, demonstrating the feasibility and advantages of 3D printed subject-specific RF coils in a clinical setting.

Methods

(i) 3D printed passive shims: Each ROI voxel of the phantom is assigned an error value (E=B_0,i,initial - B_0,i,target). A linear programming algorithm uses this spatial map of values to calculate which distribution and concentration of ferromagnetic ink are required in the shim volume to achieve the target B₀ distribution. Both software and hardware (fig. 1A) were developed to enable complete control over (ketone-based) binder and magnetic ink composition and volume in every voxel of the printed (PMMA) object. X-ray CT was used to visualize the magnetic modifier distribution in the 3D printed object (fig. 1B). B₀ maps of CuSO₄ phantoms were acquired using a 9.4T pre-clinical scanner (Bruker). (ii) 3D printed subject conform RF coils: An automated digital design workflow was developed in Python for the geometric CAD modeling of RF coils based on subject-specific input data (Python, Rhino 7). These designs were used as direct input for EM simulations (COMSOL Multiphysics) (fig. 2A). RX coils were fabricated by direct liquid metal injection of 3D printed coil template channels. The performance of these coils was compared to reference standards (16 array commercial hand/wrist coil and molded Cu wire). MRI data of CuSO₄ phantoms and the MCP joint of a volunteer were acquired (3T, Siemens). SNR maps were calculated for images formed from cov-rSoS and T1-weighted turbo spin-echo images (T1_TSE) (matrix: 384x384, TR=960, TE=15 , α=150, slice=3.3 mm, 20 images).

Results

(i) 3D prints containing magnetic functionalized ink to produce 0^th, 1^st, and 2^nd order SHs were successfully printed (xyz resolution = 300 µm) (fig. 1B). The response of the 3DP shim configurations to a B₀ field of 9.4 T was analyzed using B₀ maps (fig 1C). The 1^st and 2^nd order SHs were clearly visible in the respective B₀ maps. (ii) A phantom fitting the subject-specific thumb coil (fig. 2D) was used to quantitatively analyze the performance improvement over a commercial 16 channel volume hand/wrist coil and non-conformal Cu surface coil (fig. 2B and 2C). SNR gains of 35% for the subject-specific liquid-metal-injected coil were obtained. The non-conform copper wire coil showed a decrease of SNR of 16% compared to the 16 channel array. In vivo SNR of the thumb MCP was obtained for a subject-conform single loop Rx coil and compared to a state-of-the-art 16-channel hand/wrist array, demonstrating the feasibility of the proposed 3D printing techniques in a clinical setting. A significant increase in SNR was obtained as a result of the improved filling factor of the subject-conform coils. Ligaments and cartilage were more clearly visible compared to the reference coils.

Conclusion

The developed binder jetting 3D printing system and software technique was used to print functionalized 3D prints that induce 0th, 1st successfully, and 2nd order SH terms in B0. The results show that the proposed workflow can be used to sculpt the magnetic field distribution towards a more uniform state via the passive response of the 3DP shims to the primary B0 field. Significant SNR gains (up to 35% compared to state-of-the-art Rx coils) and enhanced spatial resolution obtained from the 3D printed subject-conform coils were achieved during the imaging of the MCP joint. These results demonstrate that 3D printed RF coils can diagnose imaging regions where standard coils don't suffice. These results demonstrate how 3D printing can be exploited as a time- and cost-efficient manufacturing method for the fabrication of subject-specific MRI hardware components.

Acknowledgements

The authors acknowledge the financial support from KU Leuven (Project No. C3/21/027 and Project No. IDN/20/016), the Research Foundation Flanders—FWO Vlaanderen (SB Scholarship No. 1SB8519N), and the European Innovation Council (Horizon 2020: GAMMA-MRI 964644, EIC Women Leadership program).