Non-invasive measurement of blood glucose by infrared, but it is far-infrared

A group of researchers, led by Professor Yuji Matsuura of Tohoku University’s Graduate School of Biomedical Engineering, has developed a method of measuring blood glucose using far infrared light.

Other researchers have proposed and developed non-invasive methods for glucose measurement using near infrared light. This method works on the premise that near infrared light of some specific wavelengths are selectively absorbed by glucose in the blood. However, accurate and stable measurement using this method has proven difficult because the near infrared light is not only weakly absorbed by glucose, but also by water, protein and haemoglobin.

In contrast, far infrared light with wavelengths of around 10 µm is strongly absorbed by glucose, making it possible, in theory, for patients to get more sensitive and accurate measurements. However, far-infrared light penetrates only a few µms from the skin’s surface, which makes the detection of blood glucose difficult. Professor Matsuura’s team has developed a new measurement technique that consists of a small prism attached to the ends of flexible hollow-optical fibres to radiate far infrared light. By using this method, it is possible to irradiate the oral mucosa of inner lips that, unlike skin, have no thick horny layer.

Results from experiments show blood glucose levels sensitively detected and accurately measured with a less than 20% margin of error, which Professor Matsuura believes is good enough for clinical uses. Diabetes is a serious problem that affects millions of people worldwide. By combining the new method with far infrared lasers that have recently been developed, Professor Matsuura expects compact and low-cost blood glucose measurement systems to soon be widely used in clinical fields.

Blog tags: 


td's picture

I think that 10μm is in the mid-ir region.

hlmark's picture

Tony - Different communities have different criteria for what's Near, Mid and Far infrared. We chemists consider Near-IR to be anything from the end of the visible to 2500 nm, mid-IR exetends from 2500 to about 50,000 nm and far-IR is anything beyond that. I once found out that on the other hand, the medical community considers "Near"-IR to go only to about 1000 nm (roughly the same as what you want to call the "Herschel" region) and "Far-IR is anything beyond that (they don't define "mid-IR at all, unless they've changed their temrinology since then). So what all these terms mean depends on who you're talking to. Let's stick to the numbers; at least 2500 nm will mean the same thing to everybody!


hlmark's picture

Ian – I have some more significant comments to your message as well. Yes, it’s a potential major breakthrough, but I have several misgivings about this.

First, the concept of a cellphone-mounted NIR spectrometer is not new. Over the past 8-10 years, at least a half-dozen to 10 proposed cellphone spectrometers have appeared in print. One of the first was a high-school science fair project. Clearly not a major backing for this, but the concept was shown. Others were more serious commercial attempts. So far, I haven’t seen any of them making inroads in the commercial sector, nor being shown at Pittcon.

Second, as a previous comment said, the performance specifications for the device are unknown. It seems highly likely to me that the previous attempts at creating such a device foundered on the hard rocks of reality; while some sort of a spectrometer can be built if you don’t care about it’s performance, it is not easy to create such a device with sufficient sensitivity, low noise, precision and stability in the face of field use. Especially if it’s going to be mounted on a cell phone. Especiallly then, sample presentation becomes a major issue; you can't just "point and click" like Spock's tricorder. Would that we could, though!

Third, who is going to calibrate it and validate the calibration,  much less maintain the calibration? Given the difficulties even experienced scientists have in calibrating the highest performance, sensitive, stable, lab-quality instruments, you certainly can’t leave that to any random self-selected user to perform the calibration step.

Fourth – the last “big thing” in the NIR arena was going to be the portable, personal NIR glucose meter. It also had all the hype that we see here, but we still are waiting to see that instrument on the market or shown at Pittcon, despite having the backing of known big companies like J & J. The laws of physics, unfortunately, are not impressed by the size of the company backing a new product.

I really don't like being such a spoilsport, but we can't let the hype over a new device blind us to the realities we have to face, too. Like the glucose meter, I fear me that the potential of this concept may remain just that: a potential. LIke so many other good ideas, the reality of an implementation of it is probably 5-10 years away - - - and may always be!