Method and system for increasing optical instrument calibration and prediction accuracy within and across different optical instrument platforms
A method and system for increasing optical instrument calibration and prediction accuracy within and across different optical instrument platforms can include two main submethods or routines. The first routine can include one for correcting differences among optical instruments of different model types regardless of the manufacturer and correcting differences among optical instruments of the same model type. The second main routine or sub-method can include one for analyzing new samples of a product over time and maintaining prediction accuracy as compared to a reference method over time. The first routine can include a “TRANS” procedure, a “MIN” procedure, and checkcell tests. The second main routine provides techniques on how a product database can be cleaned, condensed, and expanded automatically as it is used by one or more different optical instruments.
1. A computer-implemented method for increasing optical instrument calibration and prediction accuracy within and across different optical instrument platforms, comprising:
generating a MIN file that minimizes differences among optical instruments of a same optical instrument model by scanning one or more sealed product standards with a plurality of optical instruments of a same optical instrument model;
receiving data conducted from one or more tests with a checkcell;
creating a prediction file based on the MIN file for an optical instrument; and
predicting properties of a sample with an optical instrument that applies the prediction file.
2. The computer-implemented method of claim 1 , wherein the one or more tests comprise one of a checkcell optical performance test, a wavelength accuracy test, a photometric accuracy test, and a prediction repeatability test.
3. The computer-implemented method of claim 2 , wherein the checkcell optical performance test comprises developing a checkcell product file by applying a partial least squares (PLS) algorithm to an independent file comprising several scans of the checkcell.
4. The computer-implemented method of claim 2 , wherein the wavelength accuracy test comprises comparing optical wavelength peaks produced by an optical instrument scanning the checkcell against known optical wavelength peaks based on the material used in the checkcell.
5. The computer-implemented method of claim 2 , wherein the photometric accuracy test comprises verifying that absorbance of spectra in material contained by the checkcell agrees mathematically with a known form and shape of the material contained by the checkcell.
6. The computer-implemented method of claim 2 , wherein the prediction repeatability test comprises developing a prediction equation for the checkcell by using a database and applying a partial least squares regression algorithm to the database.
7. The computer-implemented method of claim 1 , further comprising storing the results of the checkcell tests and the MIN file in a location which is assessable by an optical instrument.
8. The computer-implemented method of claim 1 , wherein each optical instrument is coupled to a computer communications network.
9. The computer-implemented method of claim 1 , further comprising generating a TRANS file by comparing scans of sealed product standards made by a master instrument and an optical instrument under consideration.
10. The computer-implemented method of claim 1 , further comprising identifying a sample being scanned by the optical instrument which is outside of a product database based on Neighborhood H values.
11. The computer-implemented method of claim 10 , further comprising prompting an operator of the optical instrument to upload reference values for the sample which is outside of the product database.
12. The computer-implemented method of claim 1 , wherein generating the MIN file further comprises generating a repeatability file based upon multiple scans of one or more sealed product standards with a plurality of optical instruments of a same optical instrument model.
13. The computer-implemented method of claim 1 , wherein conducting one or more checkcell tests with a checkcell comprises conducting one or more checkcell tests with a checkcell comprising acetyl.
14. A computer-implemented method for increasing optical instrument calibration and prediction accuracy within and across different optical instrument platforms, comprising:
cleaning a product database by removing outlier product samples based on values derived from a regression equation, the product database comprising spectra and reference values; and
condensing the product database by: calculating averages of values derived from the regression equation applied to the cleaned product database, increasing the averages of the values by a select percentage, and then averaging product samples contained within limits defined by the increased averages of the values.
15. The computer-implemented method of claim 14 , further comprising predicting properties of a sample with an optical instrument that compares scans of the sample with the product database.
16. The computer-implemented method of claim 14 , further comprising separating the product database into individual constituent databases.
17. The computer-implemented method of claim 14 , wherein the regression equation comprises partial least squares.
18. A system for increasing optical instrument calibration and prediction accuracy within and across different optical instrument platforms, comprising:
a computer communications network;
a plurality of optical instruments coupled to the computer communications network; and
a computer server coupled to the computer communications network, the computer server having a processor for executing instructions that make the processor operable for:
cleaning a product database by removing outlier product samples based on values derived from a regression equation; and
condensing the product database by: calculating averages of values derived from the regression equation applied to the cleaned product database, increasing the averages of the values by a select percentage, and then averaging product samples contained within limits defined by the increased averages of the values.
19. The system of claim 18 , further comprising a project manager client coupled to the computer communications network and being able to access files for each optical instrument managed by the computer server.
20. The system of claim 18 , wherein each optical instrument automatically conducts one or more tests with a checkcell, wherein each test comprises one of a checkcell optical performance test, a wavelength accuracy test, a photometric accuracy test, and a prediction repeatability test.