Modulation-based imaging
A random or spiral mask in the beam turns any source into a wavefront sensor. One reference image, one sample image, and the full wave-optical signal falls out.
Research
One method runs through all of this: pattern the beam rather than the detector, put the physics in the forward model, and recover the picture by processing — whether what comes back is a wavefront, a scattering orientation or an elemental map.
Four ways to look at the same sample
Modulation-based imaging pulls attenuation, refraction and scattering out of a single acquisition — and on a laboratory source, not only at a synchrotron. Switch channels to see what each one reveals.
Simulated phantom — to be replaced with a real dentin or lung dataset.
Transmission. What a conventional radiograph sees: dense inclusions absorb, everything else is a soft grey. The fibrous regions are effectively invisible.
Research axes
Directional dark-field
Unresolved structure scatters anisotropically. We recover its orientation and spread per pixel — and we are pushing past the rank-2 tensor model that everyone currently assumes.
Single-pixel sensing for fluorescence and scattering
A detector with no spatial resolution at all, plus a patterned beam, plus reconstruction — and you get a fluorescence or scattering map anyway. It is how we reach contrasts that no camera can image directly.
Phase retrieval and inverse problems
Single-shot joint retrieval of phase and attenuation, variational solvers with the right regularisation, and learned priors where they genuinely beat the analytic ones.
Applications: virtual histology and heritage
Lung, heart and fetal tissue imaged without sectioning or staining — and, when the beam is free, ink and text hidden inside objects nobody is allowed to open.
Instruments
Laboratory. A Xenocs Xeuss 3.0 operating at 8.6 keV, λ = 1.44 Å, with a 75 µm pixel pitch and a variable sample-to-detector distance. It is a SAXS instrument, which means dark-field measurements can be validated against direct scattering data on the same sample without moving it.
Synchrotron. Regular campaigns at ESRF, ALBA and SOLEIL, on beamlines suited to propagation phase contrast and to monochromatic dark-field over a wide energy range.
Add the specific beamlines and the detectors, once you decide how much detail is useful.
Can you measure my sample?
Often, yes — and the honest answer usually depends on three things: how thick the sample is, whether the structure of interest is above or below the pixel, and whether it can travel. Write with a photograph and a sentence about what you are trying to see.