Title: TESTING AND VALIDATING INSTRUMENT PERFORMANCE
Munday, Cap - SYDCO, INC
Zapf, Mike - MCCORMICK, INC.
Reeves Iii, James
Submitted to: Near-Infrared Spectroscopy in Agriculture-A Tri-Societies Monograph
Publication Type: Book / Chapter
Publication Acceptance Date: December 20, 2002
Publication Date: July 1, 2004
Citation: Munday, C., Zapf, M., Reeves III, J.B. 2004. Testing and validating instrument performance. In: Roberts, C.A., Workman, J., Reeves III, J.B. editors. Near-Infrared Spectroscopy in Agriculture. Madison, Wisconsin. American Society of Agronomy. p. 33-48.
Interpretive Summary: Near-infrared spectroscopy is an instrumental method of analysis using light beyond the range of human sight. The basic concept is that the composition of a substance can be determined by how the material in question interacts with or absorbs specific wavelengths of light. In order to carry out such analysis an instrument called a spectrometer is required. Also, in order for the results to be repeatable over time, the spectrometer must produce the same results from one time to the next. In order to achieve this instrument stability and even to check that the spectrometer is performing within specifications at any time, a series of tests need to be carried out. The type and methods used for these test depend on the particular instrument configuration to some degree, but generally involve testing for wavelength accuracy (i.e., the spectrometer produces the same light output each time), a noise test (i.e., the signal strength versus background noise is consistent from time to time) and total energy output. Other tests can include testing the physical components such as sample holders and movement (i.e., rotating sample cups), that the instrument is free of sample contamination, etc. This chapter describes the types of tests which can be done, the results that should be expected, how the tests vary or are applicable to specific instrument designs and how often and by whom they should be done.
The objective of near-infrared (NIR) spectroscopy is to obtain spectra from which quantitative or qualitative calibrations for an analyte of interest can be developed. A NIR instrument can deliver inaccurate results for one of two reasons: 1. The calibration failed, or 2. The instrument failed. This chapter covers the various tests that are performed on a routine basis to assure consistent instrument performance. These routine diagnostic tests encompass spectrometer, sampling attachment and operational performance. While one-time calibrations may be possible with a poorly performing instrument, it is unlikely that such calibrations will work well with new samples in the future. The purpose of Instrument Validation is to: 1. Determine if a NIR instrument is performing according to specifications. 2. To provide the means to correct any instrument misalignment or malfunction if within the domain of the user. There are two general areas within which such tests can be classified: 1. Those based strictly on physical measurements of instrument performance. 2. Those based on the utilization of test samples and calibrations previously developed for the instrument in question. This chapter discusses the types of tests used to validate NIR spectrometers and how to interpret the results of those tests.