A new biomedical imaging system from MIT uses clever math and a $100 gaming device to make the same measurements that currently require a $100,000 piece of gear. Using the same principles that untangle cell-phone signals from the morass of radiation that fills the air around us, the team can replace expensive lab equipment with a camera from Microsoft’s Kinect console.
The team, led by MIT Media Lab grad student Ayush Bhandari, use the Kinect’s depth sensor–which usually detects your movements in a room so you can control games–to carry out fluorescence lifetime imaging, a technique that could be useful for cancer diagnosis and DNA sequencing. Biological material is treated with fluorescent dye, which absorbs light and then re-emits it a fraction of a second later. The chemicals that make up the sample interact with the dye, slowing down this re-emittance depending on their composition. By measuring this delay, you can work out what’s in the sample.
The problem lies with that fraction of a second. The detectors needed to see the delays work on the scale of picoseconds, whereas the Kinect detector only works at the nanosecond level. This is where the math comes in.
A Fourier transform is a mathematical technique that lets you take a signal and break it down into the smaller signals that make it up. Imagine tossing a handful of gravel into a still pond. The individual ripple spread out, and then bounce off each other, and before long the pond is a shivering mass of what look like random waves. The $100,000 microscope is quick enough to detect these ripples before they interfere with each other. The cheap, off-the-shelf one is too slow, and just sees the mess that’s left over. Using the Fourier transform, you can effectively run the clock back, pulling out the original waves from the mix.
It sounds like magic, but it’s at work every day, pulling our Wi-Fi streams and cell-phone conversations out of the radio soup made up of everybody else’s wireless traffic, and now the MIT team has applied it to medical processes.
While the Kinect-based rig used by Bhandari’s team doesn’t have the resolution of the expensive lab gear, it’s close enough and can be improved with more sensors. More importantly, the low cost puts this technology into the hands of anyone who wants to play with it, which is almost guaranteed to reveal new applications for the method.
As Adam Cohen, a Harvard University professor of chemistry, chemical biology, and physics told MIT News, “If I had one of these devices, what I would do is just go looking around the world at stuff. There might be all sorts of interesting things.”