For the first time, ghost imaging using electrons have been accomplished.
Optical ghost imaging using light has been previously accomplished.
Optical ghost imaging is a useful tool that can spatially resolve the characteristics of a sample using just a single-pixel detector – rather than the multipixel arrays found in digital cameras. The technique involves splitting a beam of light into a pair of correlated beams called the signal and reference beams. The signal beam strikes the sample before hitting the single-pixel detector. The reference beam goes directly to a conventional, multipixel detector. By measuring the correlation between the intensities of the beams as they hit their respective detectors, an image of the sample can be reconstructed using data from the multipixel detector, without directly imaging the sample itself.
In this new report, this technique has been accomplished using relativistic electrons. Their motivation for applying this technique using electrons is given in the text of the paper:
Potential benefits of applying ghost imaging methods to electron-based imaging systems include the possibility to minimize image acquisition time and to reduce the dose delivered to the sample and the resulting sample damage. In addition, electron ghost imaging can be useful for experimental methods (e.g. electron energy-loss spectroscopy, or cathodoluminescence) for which spatially resolved detectors either do not exist or severely increase the complexity of the setup. A special case is the growing field of time-resolved electron scattering where the use of multi-MeV, ultrashort relativistic electron sources for both imaging and diffraction has pushed temporal resolution to the ps and fs regimes. Employing structured illumination (i.e. ghost imaging) schemes on ultrashort electron beams offers the possibility to better manage the space charge effects in the electron column.
This is another opportunity for me to point out that this is a research work coming out of accelerator physics.
 S. Li et al., Phys. Rev. Lett. 121, 114801 (2018). http://www.slac.stanford.edu/pubs/slacpubs/17250/slac-pub-17314.pdf