Slashdot Mirror
PARC Builds iPod-Sized HIV Detector
MikeChino writes "Right now it's difficult, if not impossible, to quickly detect HIV in patients living in impoverished countries. That may all change soon, though — researchers at a California outfit called the Palo Alto Research Center have built an iPod-sized handheld device that can provide an immune check-up in under 10 minutes — all with a prick of the finger. With millions of people around the world without access to a full-size laboratory, PARC's device could revolutionize the detection and treatment of HIV."
It should be clarified that what PARC is announcing is a small flow cytometer used to count T-cells that express the CD4 antigen, a particular subset of T-cells that is most affected by HIV. A dropping concentration of these cells in HIV+ patients' blood is a reasonable metric for the progression of AIDS. It should also be noted that the system PARC demonstrated is not a full-fledged CD4+ T-cell counter, because you also need to measure CD3 expression to tell apart CD4+ T-cells and other white cells that express CD4, like monocytes. Traditional flow cytometers do this by using two antibodies (CD3 and CD4) labelled with different colored fluorescent dyes and measuring the intensity of each color channel using a photomultiplier tube or avalanche photodiode (these are very weak signals!). The PARC prototype is a proof-of-concept that only measures CD4 expression so would very often overestimate the true concentration of CD4+ T-cells in blood.
The interesting thing here is that they've apparently been able to do away with PMTs and APDs as detectors using a method called spatially modulated fluorescence emission. Typically a very narrowly focused laser beam (a few microns or thinner) is used to excite the fluorescent label, so you get precious few photons out of it to detect (hence the need for high-gain detectors like PMTs and APDs). The idea here is to stretch out the excitation region and use a slit pattern to help in background subtraction since you can predict the locations (in time) of the fluorescence signal as it passes by the slits allowing for integration of that signal over a longer period of time with lower background than you would otherwise be able to with a standard photodiode. This also lets you mitigate the loss of bandwidth you'd get by stretching out your detection region since you can figure out which tiny spikes in signal from the photodiode go together based on the known pattern of spacings in the slit and the timings of the peaks.
Very clever! Definitely a lot more interesting than all the cytometer-on-chip microfluidics stuff that's been thrown around since the 80's to no notable success.