| Author | Message | ||
     Bruce H. Campbell (Campclan) Moderator Username: Campclan Post Number: 84 Registered: 4-2001 |
The points made are good ones. I will expand on several thoughts I have had in this area. Firstly, NIR in biomedical research even when noninvasive requires premission from human subjects with severe oversight. This makes this usage more expensive. Animal subjects don't require permission but there is still a lot of oversight. From the approaches that have been tried, such as glucose, it appears to me that the marketing people are looking at home usage. Indeed, there is a recent report of the possible size of home usage. When this market is approached, one must have equipment that is relatively inexpensive, either to own or to rent. This means secretive research. I haven't seen many publications on the use of NIR for expensive application areas, such as in hospitals, places that can more easily afford expensive equipment. With all that said, however, there still are researchers trying to apply NIR in biomedical areas. I assume those who are open with their intermediate results are in universities or government labs. Finally, even research that is under a cloak of secrecy can be disguised when asking a question or making a comment. So why aren't there more topics on biomedical approaches. Are we too concentrated on areas that don't apply to biomedicine? | ||
| hlmark (Unregistered Guest) Unregistered guest |
Andrew - you make some good points, and there's certainly at least some truth to them. As with many analytical situations, a lot depends on the particular application. At least one application has been successful and in commercial use for a good number of years: the pulse oximeter. Other applications have indeed been less successful, your mention of blood glucose probably being the most well-known. But attempts to measure blood glucose by NIR have been around at least since 1992, when I was called in to consult on one such project, one of my first consulting clients. At one time in the mid-90's there was a talk at a FACSS conference about NIR blood glucose measurement, and the president of a company working on it said that his marketing department estimated that there were at least 60 projects world-wide attacking the problem. Another factor is that most commercial attempts at developing a biomedical analysis are not publicized until they are successful, there is usually a large amount of money riding on the outcome to let any word leak out prematurely. However, some non-commercial work has been reported; Dave Haaland has presented his work on spectroscopic blood glucose measurement at several conferences in the past, although I don't think he's still active in that area. Howard Guntherman has presented his work on the topic at several Chambersburg meetings. There have been sporadic papers in Anal. Chem. about measurement of other blood components. So some work is being reported, but you have to dig it out. As for what the bio-optical community is doing, it might simply be that they're not aware of the techniques that we use, or they are aware and don't feel as comfortable with them as they do with the more "hard" approaches that they're used to. It took the chemical/spectroscopic community a long time to become comfortable with applying multivariate mathematics to chemical analysis, too, don't forget. Or maybe they've looked at what's been done and don't think they have any "magic answers" to the problems that have been encountered and don't want to spend their time at it. Howard \o/ /_\ | ||
| V Andrew McGlone (Unregistered Guest) Unregistered guest |
Interesting question and an associated one that I think about occasionally: Why is there no �cross-correlation� at all? Not merely on a website or discussion group but anywhere really. I�ve spent a few years now working on NIR with fresh commodities like fruit and meat. Lately I�ve turned my attention to biomedical and non-invasive NIR measurement on people. It is like jumping on a whole different boat and sailing down a different river. For instance, I recently attended the Biomedical Optics Conference (BiOS 2006) at the big Photonics West conference in San Jose. Huge conference by my standards (over 1000 companies exhibited in the trade show alone). I listened to a variety of talks but didn�t see or hear a whole lot that would be considered relevant to this NIR community. There was some stuff on glucose sensing, and NIR spectroscopy got a play there, but generally spectroscopy seems to be approached quite differently. In biomedical optics, it is fundamental approaches that are too the fore, especially �hardware� based methods for resolving out scattering and absorption parameters, and there is a heavy reliance on light transport modelling or similar to help interpret data. My perspective on this is twofold. Firstly, I think that in optical terms living body tissue is far more complicated than your typical plant, foodstuff or pharmaceutical product etc. Just think pulsing blood, the matrix of various flesh and bone components and limited options in terms of sample presentation/manipulation. Secondly, I wonder if the calibration approach of standard NIR is just plain too difficult to consider when you are dealing with people (or people derived products). Easy to dial up 100�s of samples for calibration when you are dealing with plants or foodstuffs, not so easy with people. So go beyond the relatively easy robustness of simple 2-3 wavelength MLR or ratiometric modelling and then you may have gone too far for many biomedical applications. So are there grounds for �cross-correlation�? Perhaps for �in-vitro� sample analysis only, not for �in-vivo�. Just my two cents worth �� | ||
| hlmark (Unregistered Guest) Unregistered guest |
Bruce - I'd say your best shot is to find a published article in whatever area you're interested in and contact the author \o/ /_\ | ||
| Bruce H. Campbell (Campclan) Moderator Username: Campclan Post Number: 83 Registered: 4-2001 |
I don't remember any postings done from the biomedical area. Do these people have their own list server/discussion group? If they do, how do we contact them for a possible cross correlation? If they don't have one, how can we contact people to get them involved in this one? Bruce |