[jr2]

Based on polarized light. We already had this, with several limitations but now might be able to get it small enough to fit on a smart phone with less limitations AND the ability to see neural activity in real time. Star Trek medical tricorder anyone?

http://www.sciencealert.com.au/news/20142609-26244.html

[Beskargam]

Sounds like if it can detect cancer, would this also be able to do the job of an mri machine?

[Phantom Hoover]

The article's very vague but I very much doubt it.

[Dragon]

What do they mean by "cancer"? There's no such disease as "cancer". It's a group of diseases with a roughly similar working mechanism and related causes. One might cause the other, but they're very much not the same.
Humans can "see cancer" too. Skin cancer can get pretty obvious, for example. Large tumors are easily spotted, too, but I doubt they'd do anything to polarize light, what with being covered by healthy skin. Polarized light is just like ordinary light in almost every aspect, and you can see it with a sheet of polaroid or two pairs of good sunglasses. I have no idea how polarization-based detector could work, but there might be more to it (since they're obviously not using pieces of polaroid for the camera, it's possible it's not quite what I think it is).

[Nuke]

its amazing what you can do with a camera a polarization filter and opencv.

[MP-Ryan]

Cancerous cells are fundamentally different from normal, healthy cells due to the gene cascades that are activated/inactivated in the process, and clusters of them (before and after tumour formation) interact with polarized light differently than normal cells.

Polarized light is already used in several types of cancer detection (there are a number of journal articles on this that are publicly accessible, in addition to other medical sites); this type of camera appears to expand on and potentially simplify the process.

The original source of the article jr2 posted is this:  York, T et al. "Bioinspired Polarization Imaging Sensors: From Circuits and Optics to Signal Processing Algorithms and Biomedical Applications."  Proceedings of the IEEE, vol. 102, no. 10, pp 1450 - 1469, October 2014

[Dragon]

Ah, so the innovation here is "a smartphone-grade cam able to 'see' polarization" not "using polarization to detect cancer". It makes sense, it'd be strange if we didn't do it already, but with specialized medical equipment instead of consumer cameras. I suppose the differences are just too subtle to pick up with human eye and a sheet of polaroid.

[Mika]

An interesting opening I'd say

I've already seen visualization of the blood flow to head between people who have migraine and those who don't. There's significant difference between the timings of the blood flow, those without migraine have sort of equal diffusion of the blood, those who have migraine have it delayed so that the other side of the head is a bit late. This was reported in Northern Optics (2012 if memory serves), by utilizing polarization effect and facial blood vessels. It remains to be seen whether this can actually be linked to the migraine, the results were done with a relative small group of people.

Additionally, the company I'm at has also been working with visualizing skin cancer (early determination) with a different technique. So far the clinical tests have been positive for that.