How do you send light through scattering tissue and how much light do you actually need for that? One photon per pixel of the camera would be the lower limit, you might say. Remarkably, you can go much lower, researchers of UT and Caltech show in Physical Review Letters.
Although light has promising biomedical applications, for example for measuring blood circulation or tracing tumors, the depth is often limited by the heavy scattering of the tissue. How much light do you actually need? The new results of researchers of the University of Twente in The Netherlands and Caltech in Pasadena, shows that the intuitive lower limit of one photon per pixel actually is not the lower limit. Thanks to the wave character of light, even a few thousandth of a photon per pixel is sufficient. For several reasons, this is good news, as you can’t simply use more light: too much of it can damage the tissue.
Back tracing
The small amount of light that finds its way through tissue, has travelled a complex path. It is scattered many times, but eventually finds a way out. If you manage to go back along this path, you know what waveform is needed to send light through tissue with success. Although you don’t know the exact path in that case, you do know that there is a path: you calculate the result back to the source. In this way it is also possible to focus light inside tissue, enabling looking through tissue or deeper inside the brain.
One photon, multiple paths
Imagine no more than 1000 photons traveling through tissue, while the camera chip has 200.000 pixels. The first thought is that just 1000 pixels receive light, showing an occasional ‘speckle’ here and there. This is not the correct assumption, however. Different pixels can, at the same time, register the information of one single photon. As light is a wave as well, one photon can travel different paths. The phase of the light falling on the camera pixels, is always a combination of the actual signal and a reference source. Even with an ‘unequal ratio’ of pixels and photons - like 200.000 versus 1000 in the example -, the full image is available and can be calculated back to the source. Although the image has less contrast, it remains possible to reconstruct it. That is something you wouldn’t expect seeing photons as separate particles. This counterintuitive result proves that you need far less light to go deeply into tissue. This is good news for applications in new imaging techniques, for example hybrid techniques that use a combination of light and ultrasound.
‘Deep Vision’ ERC Grant
This research has been done in the Biomedical Photonic Imaging group, part of UT’s MIRA Institute for Biomedical Technology and Technical Medicine, in cooperation with electrical engineering colleagues of the California Institute of Technology (CalTech). Research was partly financed through the ‘Deep Vision’ Starting Grant of the European Research Council (ERC) that Ivo Vellekoop received.
The paper ‘Optical Phase Conjugation with Less Than a Photon per Degree of Freedom’, by M. Jang, C. Yang en I.M. Vellekoop, appeared in Physical Review Letters
