NIR accreditation

Nhi's picture

Dear all,
After we had already calibration models, do we need to get accreditation for NIR analysis? If yes, what standard we can use?
How to prove that a NIR analysis can be replace for reference method?
Thanks in advance,
p/s: used for animal feeding stuffs, cereals.

Amigdalia's picture

We have accreditation for NIR analysis for bakery products, and we made "self" reference. Bakery sample (milled, dried) can be stored - we know the reference data from this sample (we measured it earlier), and when we measure a new sample with NIR, we measure always this "self" reference too. So You can follow the difference between the reference and NIR data

hlmark's picture

Nhilt - at this point I have to say I think you're asking the wrong people. The only way to get an answer for sure is to contact one of the accrediting organizations of interest, which for feedstufs are the ones Phil recommended early on in this discussion, and get their official requirements. They will have formal standards that you'll have to meet, and they'll probably have guidelines (and other helpful information) about how to meet them. But those are the only people who can tell you for sure.

ianm's picture

Nhilt: accreditation depends at least in part, on the material to which you propose to apply NIRS. If it is a cereal or food-type in nature you can approach the American Association of Cereal Chemists International  (AACCI) that organization will require that you or one of your associated is a member of the Association. Approval (accreditation)  will require a ring test, the results of which will have to receive approval from the AACCI Approved Methods committee, chair of which is Dr. Anne Bridges 651 454 7250 Ext 156. Anne is good people. If you are working with any other material you can approach the Association of Analytical Communities (AOAC). They will also require a ring test and approval by their membership, but work with more materials. Contacts are 1 301 924 7077; Fax 1 301 924 7089; e-mail [email protected]. If your organization is big enough certification may not be essential if you want to work in-house. You can monitor your calibrations using certified reference analytical methods.

Hope that this helps.

(posted by Ian Michael on behalf of Phil Williams)

hlmark's picture

Nhi - you didn't say what your application is, which makes it somewhat difficult to figure out what sort of "accreditation" you need. If your application is in the agricultural/bakery fields, you should follow Phil's advice. If in another field you could contact the appropriate industry's certifying organization(s). In general, though, you could expect to conform to good analytical practice; one example that I happen to be a bit familiar with is the rquirements of the FDA for analysis of pharmaceuticals. In one sense their requriements are very general, they say that you have to "demonstrate, by laboratory experiments, that your analytical method is fit for the purpose" it is intended for. However, specifics are given in other documents, including standards and practices written by other industrial, official, and independent scientific agencies. These other requireements include:
1) Precision (which is further brokedn down into Repeatability, Reproducibility and Intermediate Precision
2) Detection limit
3) Quantitation limit
4) Specificity (demonstrate that your method is in fact measuring the intended analyte and not some other component of the samples)
5) Accuracy
6) Robustness
7) Ruggedness
8) Linearity
9) The Range over which the above characterstics are within the values specified by your method.
Those are all standard parameters for chemical analysis. It may be that in your specific case you may need to include other types of tests. It may also not be necessary to test all the above charactreristics on all types of samples; but you need to demonstrate that if a sample passes all the tests you do specify, then that sample is suitable for its intended purpose, which after all is the goal of all these types of QC tests.

Ahmed's picture

Dear Nhi-
In the pharmaceutical field, NIR technology has been approved by the new pharmacopeias such as BP2016 to be used as an analytical method, however, due to the different nature of this technology from the usually used technologies, it has a different validation tests other than usually applied for chromatographic mehtods.
You can find the chapter from BP2016 concerning NIR technolgy, its applications and the validation procedure.
Along very long years of experience, I have made various NIR chemometric models for the pharmaceutical API assay, water content and even physical properties of the samples, just contact me if need any kind of support.

Ahmed Ramadan

Natural Wellness Egypt Scientific Office  | Technical Project Assistant Manager


Nhi's picture

Thanks for all kind advices,
I want to validate the calibration models in raw material (rice, wheat, barley, blood meal, fish meal...) and animal feed (shrimp, pet, pig, poultry...). I'm reading ISO 12099 and ISO 21543, it is the tool to validate products in animal feeding stuff, cereals and milled cereals, but I don't know it's enough to get certification for NIR analysis or not?
Best Regards,