Revealing the invisible through wave signatures

Finding a needle in a haystack? An object hidden in thick fog? Conventional imaging techniques, whether ultrasound, microscopy, or radar, all face the same obstacle: multiple scattering. When a wave propagates through a highly heterogeneous medium, it is scattered numerous times before returning to the detector. The resulting image becomes blurred, to the point of completely obscuring the target being sought.

To overcome this limitation, researchers have developed an innovative approach based on a simple yet effective concept: the matrix fingerprint. Every object has a unique signature, its own distinct way of scattering waves. This signature can be described mathematically by a reflection matrix, which serves as the object’s “fingerprint.” Even when buried in a highly scattering medium, the object retains its fingerprint; it is merely drowned out by ambient scattering noise.

The researchers’ idea [1] is to exploit the correlations between the matrix measured in the fog and the reference pattern. These correlations make it possible to isolate the specific contributions of the target and make them visible with a precision better than the wavelength used. Instead of correcting for the effects of noise, a virtually impossible task in opaque media, the method reveals what in the signal is resistant to interference. To validate this principle, the team conducted several experiments in ultrasonic acoustics. Metal spheres buried in a granular suspension, which are completely invisible using conventional ultrasound, could be located with near-certain reliability. In a second test, the method detected lesion markers used in breast cancer diagnosis, which are often difficult to distinguish amid tissue noise. Finally, when applied in vivo to a human calf, it revealed the architecture of muscle fibers, paving the way for quantitative tissue imaging, which is invaluable for diagnosing cardiac or neuromuscular diseases.

Beyond these proof-of-concept demonstrations, the matrix fingerprint stands out for its flexibility and versatility. Applicable to any type of wave, provided a multi-element sensor is available, it paves the way for a wide range of applications: monitoring medical instruments in interventional radiology, non-destructive testing of materials, improving radar and sonar systems, and even local measurement of physical parameters such as temperature or pressure, which are typically inaccessible from a distance.

By revealing what lies hidden in the fog of electromagnetic waves, this approach offers a truly new way of seeing the invisible, with perspectives that extend far beyond the realm of fundamental physics.

This research was supported by an ERC Consolidator Grant (No. 819261) under the European Union’s Horizon 2020 research and innovation program. This innovation is also the subject of a patent filed by SuperSonic Imagine, jointly owned with the CNRS and ESPCI (French Patent No. 2314789, 2023).