Research & Papers

Optical detuning slashes MRI coil power to 10 mW for flexible arrays

Photodiode-PIN combo matches galvanic performance at a fraction of the power

Deep Dive

MRI receive coils require detuning during transmit to avoid excessive currents and patient safety risks. Traditional galvanic detuning uses wired connections that limit coil flexibility and increase complexity. In a new arXiv preprint (arXiv:2607.06849), Jakob Gerlach and colleagues from the University of Freiburg systematically compare four optical detuning strategies for shielded loop resonators (SLRs) at 3T, aiming to replace wires with light-based control.

The team tested photodiode-PIN diode combinations, phototransistors, and passive detuning against conventional galvanic detuning. The photodiode-PIN diode combo performed best, achieving an unloaded Q factor of 68.6 and a Q ratio of 1.9 at just 10 mW optical power — performance comparable to galvanic detuning in bench tests and phantom SNR measurements. A flexible four-channel SLR array using this strategy was built and validated in vivo, producing high-quality knee and brain images.

The shift to optical detuning promises fully flexible, lightweight coil arrays that eliminate electrical cables, reducing patient discomfort and enabling new MRI applications. While passive detuning (wireless) also worked, active optical control provides robust detuning across all channels. The authors note that optimizing optical power requirements — already modest at 10 mW per channel — is a critical step toward clinical adoption of all-optical receive arrays.

Key Points
  • Photodiode-PIN diode combo achieved unloaded Q of 68.6 and Q ratio 1.9, matching galvanic detuning
  • Total optical power requirement of 10 mW per channel — comparable to conventional wired solutions
  • Four-channel flexible SLR array demonstrated successful in vivo knee and brain imaging at 3T

Why It Matters

Optical detuning paves the way for cable-free, flexible MRI coils, improving patient comfort and image quality.

📬 Get the top 10 AI stories daily