Shining a light on diabetes
Aug 18, 2010 4 comments
A quick and painless way to measure blood sugar is highly sought-after by diabetes sufferers, who currently have to prick their fingers to draw blood several times a day. Now, researchers in the US may have found a solution – a device that works by simply shining a light on skin.
The vision is to create a laptop-sized device that could be kept at home or carried around. Rather than having to pierce the skin to obtain blood samples, the device measures sugar levels by simply placing a scanner against the skin. Because measurement is fast and easy, it is hoped that the device may encourage people with diabetes to check their blood sugar more often, giving them better control over their condition.
At the heart of the device is a Raman spectrometer, which can identify chemical compounds by measuring how near-infrared laser light scatters on contact with molecules. The idea of using Raman spectroscopy to measure sugar levels in blood was first suggested 15 years ago by Michael Feld at the Massachusetts Institute of Technology (MIT). Although Feld sadly passed away in April this year, his team is now starting to realize his vision.
Keeping up with the sugar rush
The problem until now has been that near-infrared light can only penetrate a short distance into the skin. The technique therefore detects glucose in the fluid surrounding skin cells (the interstitial fluid), rather than in the bloodstream. This is a problem because blood glucose levels can change rapidly, such as after eating, while there is a time lag of 5–10 minutes before the sugar changes can be seen in the interstitial fluid.
The MIT team has now resolved the problem by developing an algorithm to relate blood glucose to interstitial glucose levels. "We’ve incorporated a mass-transfer model into the overall Raman spectroscopic algorithm, which allows us to seamlessly transform between blood and interstitial fluid glucose," explains Ishan Barman, lead author of the research.
Using an early version of the device, the team tested the blood-sugar levels of some human volunteers and found that the accuracy and precision of the test was just as good as conventional finger-prick tests. In addition, the new algorithm allows the test to predict impending episodes of high or low blood sugar (hyperglycemia and hypoglycemia) by extrapolating the rate of change of sugar concentration.
The next challenge is to strip down the Raman system and build a miniaturized device that would be suitable for home use. A prototype has been made and is already scheduled for clinical testing, but reducing the complexity of the system and shrinking down bulky components could take a while.
"We are in a proof-of-concept stage in terms of device development – and we envision a laptop-sized or hand-held unit that could cost as little as $200," he told physicsworld.com. "It is difficult to predict due to market variations and FDA regulations, but one could anticipate an optical device for glucose monitoring in the next 5–7 years."
Randall Jean, an expert in remote sensing at Baylor University in Texas, US, is impressed by the work. "This research addresses a real problem and appears to provide an important means for improving the calibration of non-invasive sensors," he says. "It may also be helpful in the development of a so-called 'artificial pancreas' – where insulin can be dispensed automatically in response to sugar levels."
This research is described in Analytical Chemistry.
About the author
Lewis Brindley is a science writer based in Reading, UK.
I Barman et al., Anal. Chem., 2010. DOI: 10.1021/ac100810e