Method for creating spectral instrument variation tolerance in calibration algorithms

• US 6,615,151 B1
• Filed: 08/28/2001
• Issued: 09/02/2003
• Est. Priority Date: 08/28/2000
• Status: Expired due to Term

• Current Assignee: Covidien LP (Medtronic Plc)
• Original Assignee: CME Telemetrix, Inc.
• Inventors: Scecina, Thomas, Cadell, Theodore E., Pawluczyk, Romuald
• Primary Examiner(s): Barlow, John
• Assistant Examiner(s): LE, JOHN H
• Application Number:
  US09/939,707
• Time in Patent Office: 735 Days
• Field of Search: 702/28, 702/85, 702/86, 702/90, 702/127, 703/2, 250/573, 600/322, 700/110
• US Class Current: 702/85
• CPC Class Codes: G01N 21/274 Calibration, base line adjustment, drift correction
• CPC Class Codes: G01N 21/359 using near infrared light

Abstract

The present invention relates to a method of building instrument variation tolerance into calibration algorithms for spectroscopic devices for chemical composition analysis with spectroscopic methods. The method of the present invention is particularly suitable for blood glucose, cholesterol and other chemical components prediction based on near-infrared spectrophotometry measurements. A method includes developing a calibration algorithm on a first instrument; applying the calibration algorithm to a second instrument; calibrating the second instrument and adjusting the calibration algorithm to account for differences between the first instrument and the second instrument; and repeating the step of applying (above) in respect of (n) further instruments to provide a calibration algorithm which may then be used on other instruments. Methods are also provided that develop calibration algorithms by adding instrument variations, based on mathematical manipulation, of the spectral data collected on calibration instruments.

Claims

1. A method of generating a calibration algorithm adaptable to instrument-specific variations in assigned wavelength, comprising:

measuring a spectral response for each one of a set of calibration samples, each measured spectral response comprising at least one measurement element, each measurement element having an assigned wavelength;

shifting the assigned wavelength of at least one of said at least one measurement element of each measured spectral response by at least one of a random amount, a uniform amount, and a variable amount, that is dependent on the magnitude of the assigned wavelength, wherein at least one modified spectral response, that differs in the amount of shifting, is generated for each of said samples; and

using a multivariate analysis method to generate said calibration algorithm from the measured spectral responses and the modified spectral responses of said samples.

2. The method of claim 1, wherein the spectral response for each of the calibration samples is measured on at least two spectral instruments, wherein at least two modified spectral responses are generated from each of the spectral responses measured on said at least two spectral instruments.