Recipe for low-cost medical tests: FISH and chips

Step by step
Performing a complicated medical test on a microfluidic chip involves metering minute quantities of a patient sample down channels and into reaction chambers where reagents are pumped. Typically, several such tests must be performed on different chambers on the chip to obtain final test results. Detectors then determine the outcome of the test and what should be done next.

"Not only do you need to use the smallest quantities possible to speed up results and to make them cheaper, but you also need quite a bit of intelligence in your microfluidic device, because the procedures are very complex," said Backhouse. "You need micropumps, microvalves, microreservoirs and microdetectors, all automated and under computer control."

Today many medical laboratories are only set up to perform such tests manually, requiring a skilled technician to perform all the steps. Equipment is available to automate such tests, using robots to perform the steps of the technician. But such machines can cost up to $1 million each; can only operate in batch operations, in which many other identical tests are performed simultaneously on different samples; and can sometimes take days to get the final results.

With a microfluidic device, both the cost and time are cut by virtue of the small scale of the test. The smaller the scale, the less reagent needed. In addition, such tests often depend on diffusion over measurable distances; thus, the smaller the scale, the shorter the diffusion distance and the faster the test results can be obtained.

So far, the microfluidic chips have automated the most tedious, time-consuming and labor-intensive elements of the tests. Before commercial versions of the experimental microfluidic chips can be marketed, however, the researchers plan to automate the more-esoteric elements of the tests, so that entire batteries of diagnostic procedures can be performed while the patient waits.

"We want physicians to be able to ask a question and get the answer quickly. The classic line is this: 'We want the surgeon to know when to stop cutting, rather than make a guess, wait and see, then operate again if their guess was wrong,' " said Backhouse. "So far, we have integrated the more arduous steps that are well established onto our microfluidic chips, but we still need an expert in a white coat to supervise its operation."

The team hopes to design a completely automated, CMOS-compatible microfluidic chip that is inexpensive enough to be sold in pharmacies and that does not require trained personnel, similar to a home-pregnancy test. Such end-user-operable chips would level the playing field, so that everyone, not just the wealthy, could have access to the most modern medical diagnostics available.