Lorentza images the bio-electric layer of tissue — its electrical conductivity and permittivity — by fusing focused ultrasound with room-temperature quantum magnetometry. Functional contrast that appears weeks to months before a mass is visible. Radiation-free, unshielded, portable.
Ultrasound localises the signal; the quantum sensor reads it. Spatial resolution is set by sound, sensitivity by quantum optics.
Every clinical imaging modality shares one assumption — that pathology becomes detectable when it becomes morphologically visible. But malignancy and internal bleeding are biophysically active long before they form a mass. They alter ion transport and membrane integrity first, shifting tissue conductivity (σ) and permittivity (ε). No portable tool reads that layer today.
sees · X-ray attenuation
Ionising; misses up to 30% of cancers in dense breasts
sees · Proton density & relaxation
Structural, not functional; heavy, shielded infrastructure
sees · Acoustic impedance
Reflects boundaries, blind to bio-electric state
sees · Metabolic tracers
Ionising radiotracers; poor spatial resolution
Breast-cancer 5-year survival is 99% when caught localised — and falls to 27% once it has spread.
The single biggest lever in survival is detecting disease earlier. Lorentza opens a new contrast channel that activates before the structural one.
Classical magnetometry is trapped by a trade-off: shrink the sensor for resolution and you destroy its sensitivity. Lorentza breaks it by delegating spatial selection to the acoustic focus — so a large, ultra-sensitive sensor can read a signal it knows came from a millimetre-scale spot.
A programmable phased array focuses ultrasound (1–10 MHz) into a sub-millimetre voxel deep inside tissue — the same wavefront-shaping we pioneered for light, applied to sound.
Inside a compact 0.1 T bias field, the oscillating ions feel a Lorentz force. That drives an RF current density whose amplitude and phase encode the complex conductivity σ̃.
The current radiates a tiny RF magnetic field (~1 pT). A room-temperature ⁸⁷Rb optically-pumped magnetometer detects it — no cryogenics, no superconductors.
Homodyne detection phase-locked to the ultrasound clock splits the signal: the in-phase channel gives conductivity σ, the quadrature channel gives permittivity ε — co-registered, in one acquisition.
The observable is not a scalar image but the complex conductivity of tissue — a direct physical signature of membrane integrity and water compartmentation.
Focused ultrasound imposes a particle-velocity field. In a static bias field $\mathbf{B}_0$, the Lorentz force drives an RF current density whose complex amplitude carries both conductivity and permittivity:
Demodulating the measured field $B_{RF}(t)$ synchronously at the drive frequency $\omega_{ac}$ yields two orthogonal channels — a single acquisition produces co-registered σ and ε maps:
in-phase · conductivity
quadrature · permittivity
A back-of-the-envelope chain shows the signal sits squarely within reach of an optimised warm-vapor sensor — the engineering target, not a miracle.
Acoustic pressure
clinically safe focal amplitude
100 kPa
Particle velocity
p_{ac}/(ρ cₛ), soft tissue
≈ 0.07 m/s
Current density
σ ≈ 0.8 S/m, B₀ = 0.1 T
≈ 5.6 mA/m²
RF field at sensor
3 cm from 1 mm³ voxel
≈ 1 pT
Sensor floor
warm ⁸⁷Rb OPM
≤ 100 fT/√Hz
The expected ~1 pT signal exceeds a ≤100 fT/√Hz sensor floor with margin for synchronous averaging.
Biomagnetic imaging (MEG, MCG) works at DC–kHz, where environmental 1/f noise reaches ~10⁻⁷ T and demands magnetically shielded rooms costing more than the instrument. By modulating the biological signal at the ultrasound frequency, Lorentza shifts detection into a band where the ambient magnetic noise floor drops by roughly eight orders of magnitude — and phase-locked detection acts as a narrowband filter on top.
→ Femto-to-pico-tesla detection, unshielded.
10⁸×
lower magnetic noise vs DC band
0 T
shielding & cryogenics required
Lorentza is a differentiated descendant of magneto-acousto-electrical tomography (MAET) — same proven Lorentz mechanism, but with quantum transduction, complex-conductivity spectroscopy, structured excitation, and MHz quiet-window operation.
| Capability | EIT | MREIT | Classical MAET | Lorentza |
|---|---|---|---|---|
| Spatial resolution at depth | > 1 cm | ~ mm | ~ mm | 0.2–1 mm |
| Sensitivity floor | electrode-limited | high (MRI) | ~10 mV noise | ≤ 100 fT/√Hz |
| Magnet | none | 1.5–3 T SC | 1.0–1.5 T SC | 0.1 T permanent |
| Shielded room | no | yes | often | no |
| Separates σ and ε | partial | no | σ only | σ + ε in one shot |
| Electrode contact | required | no | often | contactless |
| Ionising radiation | no | no | no | no |
| Portable | yes | no | no | target: yes |
EIT — Electrical Impedance Tomography · MREIT — Magnetic Resonance EIT · MAET — Magneto-Acousto-Electrical Tomography · SC — superconducting.
A radiation-free functional channel addressing two high-priority needs at once — earlier oncology screening and rapid emergency triage. The global oncology-diagnostics market alone exceeds €30 billion.
Malignant tissue shows σ ≈ 0.82 S/m versus ≈ 0.026 S/m for fat — a 32:1 contrast at 1 MHz that grows with aggressiveness. Non-ionising and non-compressive, it works where mammography fails: the ~40% of women with dense breast tissue.
Blood is highly conductive (σ ≈ 0.82 S/m) against surrounding tissue. Portable, unshielded operation enables rapid detection of non-compressible internal bleeding and pneumothorax where MRI is unavailable.
Because conductivity tracks tumour progression, the same dielectric fingerprint offers a route to stratify lesions and follow response to therapy — a functional read-out, not just an anatomical snapshot.
Built directly on the ERC Consolidator project MISTiQ-Light — our atomic-coherence control, nonlinear optics and quantum-measurement toolkit transfer, architecturally unchanged, to ultra-weak RF magnetic detection.
Engineer the acoustic–magnetometer geometry and the lightweight Halbach permanent-magnet bias field. First Lorentz-current signatures in controlled samples.
Synchronous optical pumping and free-spin-precession readout in the MHz band; phase-locked homodyne extraction; narrowband rejection of external RF backgrounds — unshielded.
Validate sensitivity, spatial selectivity and σ/ε reconstruction in heterogeneous phantoms representative of breast tissue and haemorrhage, with a physics-informed inversion engine.
European provisional patent across the sensing geometry, pulse-compression signal chain and FNO/PINN reconstruction. Dual-track: university spin-out plus Tier-1 licensing.
Multi-modal sensing geometry
Co-registered atomic magnetometer + phased-array probe with phase-locked homodyne detection.
Pulse-compression signal chain
Barker/Golay code sequences optimised for low-field RF magnetic detection in tissue.
Physics-informed reconstruction
Fourier Neural Operator constrained by Maxwell + acoustic laws — no unphysical solutions.
We're talking to clinical partners, deep-tech investors and medical-device manufacturers as we move from proof of concept toward a portable, radiation-free diagnostic platform.