Hardware that does not exist in any catalogue.

At 7 T, conventional whole-body gradients are constrained by peripheral nerve stimulation limits and by the acoustic noise of high slew rates. A unipolar insert gradient placed close to the head can reach the field strengths needed for high-resolution imaging — but the conductor geometry must resolve the encoding ambiguity inherent to unipolar operation, while structural integration and thermal management must survive a full duty cycle inside a high-field bore.

I develop the parametric model and conductor geometry; my team at Futura handles composite layup, conductor winding, and sub-system integration through to on-site installation at ETH Zürich.

Receive-coil SNR is maximized by close, conformal contact with the patient. Rigid coil housings cannot achieve this across anatomical variation. The challenge is building a structure that is mechanically flexible enough to conform — yet dimensionally stable enough to maintain coil geometry, overlap regions and decoupling — while surviving the MRI environment and repeated clinical use.

The granted patent describes the resulting receive coil arrangement: the annotated figure on the left, the assembled prototype on the right. The translation between the two — reference numerals to a part you can hand to a clinician — is the work.

A local breast gradient conformal to bilateral anatomy must achieve the gradient strengths needed for diffusion encoding and supersonic readout while fitting within bore geometry and integrating with receive-coil structures. The conductor winding path must be derived from — and manufacturable on — a doubly-curved surface.

This is a problem that sits squarely at the boundary of electromagnetic optimization and composite fabrication. I define the geometry and manufacturing constraints; my team at Futura is realizing the industrial prototype, to be presented at ISMRM 2026.