Spectroscopy and Chemometrics News Weekly #7, 2021

NIR Calibration-Model Services

Spectroscopy and Chemometrics News Weekly 6, 2021 | NIRS NIR Spectroscopy MachineLearning Spectrometer Spectrometric Analytical Chemistry Chemical Analysis Lab Labs Laboratories Laboratory Software IoT Sensors QA QC Testing Quality LINK

Spektroskopie und Chemometrie Neuigkeiten Wöchentlich 6, 2021 | NIRS NIR Spektroskopie MachineLearning Spektrometer IoT Sensor Nahinfrarot Chemie Analytik Analysengeräte Analysentechnik Analysemethode Nahinfrarotspektroskopie Laboranalyse LINK

Spettroscopia e Chemiometria Weekly News 6, 2021 | NIRS NIR Spettroscopia MachineLearning analisi chimica Spettrale Spettrometro Chem IoT Sensore Attrezzatura analitica Laboratorio analisi prova qualità Analysesystem QualityControl LINK

This week’s NIR news Weekly is sponsored by Your-Company-Name-Here – NIR-spectrometers. Check out their product page … link

Get the Spectroscopy and Chemometrics News Weekly in real time on Twitter @ CalibModel and follow us.




Near-Infrared Spectroscopy (NIRS)

“Evaluation of swelling properties and drug release from mechanochemical pre-gelatinized glutinous rice starch matrix tablets by near infrared spectroscopy” LINK

” Multivariate method for prediction of fumonisins B1 and B2 and zearalenone in Brazilian maize using Near Infrared Spectroscopy (NIR)” | LINK

“Non-destructive determination of fatty acid composition of in-shell and shelled almonds using handheld NIRS sensors” LINK

“Breakthrough instruments and products: Near infrared spectral sensing: Advances in portable instrumentation and implementations” LINK

“Quali-quantitative monitoring of chemocatalytic cellulose conversion into lactic acid by FT-NIR spectroscopy.” LINK

“NIR Spectroscopy Detects Chlorpyrifos-Methyl Pesticide Residue in Rough, Brown, and Milled Rice” LINK

“ANALYSIS OF TEA LEAVES WITH DIFFERENT OXIDATION STATES BY FT-NIR SPECTROSCOPY” LINK




Infrared Spectroscopy (IR) and Near-Infrared Spectroscopy (NIR)

“Identification of peach and apricot kernels for traditional Chinese medicines using near-infrared spectroscopy” LINK

“Breakthrough instruments and products: Near infrared spectral sensing: Advances in portable instrumentation and implementations” LINK

“Two-dimensional moisture content and size measurement of pharmaceutical granules after fluid bed drying using near-infrared chemical imaging.” LINK

“Influence of steroids on hydrogen bonds in membranes assessed by near infrared spectroscopy” LINK

“Determination of the oxidative stability of biodiesel fuels by near-infrared spectroscopy” LINK

“In Vitro Spectroscopy-Based Profiling of Urothelial Carcinoma: A Fourier Transform Infrared and Raman Imaging Study” LINK

“Novel alternative use of near-infrared spectroscopy to indirectly forecast 3D printability of purple sweet potato pastes” LINK

“Shortwave infrared hyperspectral imaging system coupled with multivariable method for TVB-N measurement in pork” LINK




Chemometrics and Machine Learning

“Discrimination between wild-grown and cultivated Gastrodia elata by near-infrared spectroscopy and chemometrics” LINK

“A Single Model to Monitor Multistep Craft Beer Manufacturing using Near Infrared Spectroscopy and Chemometrics” LINK

“Noninvasive Blood Glucose sensing by Near-Infrared Spectroscopy based on PLSR Combines SAE Deep Neural Network Approach” LINK

“Comparison of a Low-cost Prototype Optical Sensor with Three Commercial Systems in Predicting Water and Nutrient Contents of Turfgrass: Prediction performance of …” LINK

“Development of NIR-HSI and chemometrics process analytical technology for drying of beef jerky” LINK

“Soil spectroscopy with the use of chemometrics, machine learning and pre-processing techniques in soil diagnosis: Recent advances-A review” LINK

“Application of NIR handheld transmission spectroscopy and chemometrics to assess the quality of locally produced antimalarial medicines in the Democratic Republic of Congo” LINK




Facts

“An Accuracy Improvement Method Based on Multi-Source Information Fusion and Deep Learning for TSSC and Water Content Nondestructive Detection in “Luogang” Orange” LINK




Research on Spectroscopy

“Nondestructive methods for determining the firmness of apple fruit flesh” LINK




Equipment for Spectroscopy

“Development of Fluorescence Imaging Technique to Detect Fresh-Cut Food Organic Residue on Processing Equipment Surface” LINK




Process Control and NIR Sensors

“Acid number, viscosity and end-point detection in a multiphase high temperature polymerisation process using an online miniaturised MEMS Fabry-Pérot interferometer.” | LINK

“Real-Time Monitoring of Yogurt Fermentation Process by Aquaphotomics Near-Infrared Spectroscopy.” LINK




Agriculture NIR-Spectroscopy Usage

“Rapid and nondestructive determination of qualities in vacuum packaged catfish (Clarias leather) fillets during slurry ice storage” LINK

“Economic and chemometric studies to supplement food-grade soybean variety development in the Mid-Atlantic region” LINK

“Online Monitoring of Fermented Grains Parameters for Chinese Liquor Brewing Based on Near Infrared Spectroscopy” LINK




Horticulture NIR-Spectroscopy Applications

“Fast, simultaneous and contactless assessment of intact mango fruit by means of near infrared spectroscopy [J]” LINK

“The development of portable detector for apples soluble solids content based on visible and near infrared spectrum.” LINK

“Non-destructive and fast method of mapping the distribution of the soluble solids content and pH in kiwifruit using object rotation near-infrared hyperspectral imaging …” LINK




Food & Feed Industry NIR Usage

“A two-tiered system of analysis to tackle rice fraud: The Indian Basmati study” LINK





.

NIR-Predictor – Frequently Asked Questions (FAQ)


NIR-Predictor – FAQ

Please also refer to the NIR-Predictor – Manual and check the Hints and Notes.


How to Configure / Load / Import / Activate / Setup / Use the Calibrations (*.cm) in NIR-Predictor?

Chapter “Configure the Calibrations for prediction usage” – NIR-Predictor – Manual


Do you have a calibration file for XY ?

We create custom calibrations out of your NIR + Lab measurements of XY.
We do not sell off-the-shelf calibrations.


I have downloaded the software but can’t see it?

Please note, that the download time will be very short, because of the small file size.
Check your browser’s download folder. The download is the file “NIR-PredictorVx.y.zip”


Why a .zip and no installer (Setup.exe) ?

Because a .zip deploy keeps it simple for all:

  • easy : no Administrator rights needed to install, delete it to uninstall
  • harmless : no system changes during setup
  • transparency : you see what you get

Is there a command line (CLI) version of NIR-Predictor ?

If you want to customize it in all details, our OEM API for NIR-instrument-software (White-Label) integration gives you full access. If you are an NIR-Vendor (or similar) please contact us via email info@CalibrationModel.com


The free NIR-Predictor does not create a model, what is wrong ?

As the name says, the NIR-Predictor just predicts NIR data with a model. To create a model you need to send your data to the CalibrationModel service, after development process you get an email with a link to the calibration where it can be purchased and downloaded.


Why does the creation of the PropertyFile.txt take so long with hundrets of spectra files?

It normally takes only 1-10 seconds not minutes.
Make sure that the spectra data files are stored locally on your main drive
and not on a cloud-drive or network storage or slow USB thumb drive or SD-Card.


Is there a way to use converted ASCII spectrum data to be used in NIR-Predictor?

Yes, this is the simplest ASCII CSV file format the NIR-Predictor supports.
And there are other formats supported.


Can you convert old calibration data from vendor A to be integrated to our new vendor B NIRS calibration data in our instrument?

No, we don’t do model or spectra conversion / transformation (aka model transfer).
We build optimized models with wavelength compatible data.


Does NIR-Predictor contain any malware, spyware or adware?

No, NIR-Predictor does not contain any malware, spyware or adware.


How to copy the prediction results from the table in the browser?

Copy selected columns from the table.
By holding down the Ctrl key, rectangular areas in the table can be selected with the mouse and copied to the clipboard with Ctrl+C and then copied to a spreadsheet program with Ctrl+V.


Will an expired calibration still work?

No. Until you extend the usage time.
The expired calibration file will be moved to the CalibrationExpired folder on the next start of NIR-Predictor or “Search and load Calibrations” menu function.

Selected Calibration files from the folder CalibrationExpired can be send to info@CalibrationModel.com with your Request file (.req) files for extending their usage time.

There is the possibility to get a perpetual usage, which means their is no expiration (valid until 2050).

That way you can get the time extended calibration back, that behaves exactly as the one before with extended usage time.


What does the calibration Expiration date mean exactly on the Prediction Report?

The Expiration date, it is the final day when the calibration will be valid (similar to credit cards).


What does the (number) in brackets after the [range] mean?

During the creation of a Calibration Request the NIR-Predictor shows a message containing
” ‘Prop1’ / “Quantitative [1.40 – 2.90] (154) ”
here 154 is the number of unique values in the property range.


How long does it take to create of the property file and calibration request file from 500 spectra files?

Use a local folder on your computer (not a network drive) for your spectra files then the property file and the calibration request file is created in around 1-3 seconds. (measured on a system with SSD drive, Intel i7, 2.4 GHz)


I tried to create the calibration request and got the error message: The number of Property Values of all spectra are different?

Use the generated PropertiesBySpectra (Note: Spectra not Sample) template file.
Do not reformat the generated template file, just fill in the Property values and save as Text CSV (*.csv) file (not as Excel file “.xlsx”).


Are you able to create the calibration if we only have 20 spectra?

To create a reliable quantitative calibration you need measured spectra of at least 48..60 (more is better) different samples with different Lab-values in the required measuring range!
The NIR-Predictor will check for that automatically.


Same sample measured multiple times (replicate measurements), how to enter the Lab-value only once in Property file?

Use the created PropertiesBySample template (not the PropertiesBySpectra).


Different samples with exact the same Lab-value, how to enter the Lab-Values?

Entering the Lab-values into the generated PropertiesBySpectra template the NIR-Predictor detects them as the same Sample and as a result we have too less different values to build a calibration. But in my case these are different samples with exact the same Lab-value, how to do?

You have to cheat a little bit to make NIR-Predictor do not detect same sample as measured multiple times. This is because most NIR users do replicate measurements on the same sample and NIR-Predictor looks for that. In PropertiesBySpectra modify the values a little bit to make them different, e.g.
0.18
0.18001
0.18002


Is the free NIR-Predictor software you provide for anyone to download and use?

Yes, if downloaded directly from our homepage by the user.
See also Software License Agreement


What is an Outlier?

The spectrum is an outlier to the model, if the spectrum is not similar with the spectra and lab-values the model is built with.

The default setting in NIR-Predictor Menu > “Report with Simplified Outlier Symbols”
is ON, that will show only the worst case instead of all combinations to have a simplified minimal information.
if OFF, that will show the combinations (e.g. “XO=” or “XO” or “O=” or “X=”), which is more informative for analyzing problem cases.
See also manual chapter Outliers.


If something is wrong, please tell us

Frequently Asked Questions (FAQ)

NIR-Predictor Software


Q: Where are the FAQs of the NIR-Predictor software?
A: The NIR-Predictor FAQ can be found here NIR-Predictor FAQ.


Applications


Q: How much is your service (NIR Calibration Models)?
A: Please see Pricing

Q: How many samples do I need to send you for a new calibration?
A: Data samples from 100 – 120 different products to start, at least 60 – 100.

Q: What kind of NIR application you have and what equipment they support?
A: All major NIR and FT-NIR spectrometers and mobile NIR instruments are supported. We deliver the optimal calibration settings for your custom NIR data for your applicative purpose. We do not sell any NIR spectra files or any application/calibration files.

Q: What is the maximal number of parameters of a NIR Calibration?
A: We build and optimize each NIR Calibration for one parameter. That gives the possibility to use all advantages of spectral pre-processing and variable selection to get a robust and precise model. Your data files can contain multiple spectra with multiple parameters (there is no limit).

Q: Is qualification possible with quantitative calibration?
A: For A or B qualification define a product property e.g. name it “quality” where good is 1.0 and bad is 0.0. The quantitative calibration than predicts values around the range 0.0 to 1.0 and that classify the “quality”. A,B or C could also be possible to 0.0 0.5 1.0. Add multiple properties to your samples, like “qualityAB” and “qualityABC” and we try all the properties to calibrate.

Compatibility


Q: Are our NIR Calibration Models independent of any kind of spectrometer?
A: The calibrations we develop with your NIR data are customized for your spectrometer (any brand).

Q: Are the NIR Calibrations compatible with my NIR spectrometer?
A: The customized calibrations based on your data will fit your NIR spectrometer’s resolution and wavelength range.

Q: What do you mean by “to rebuild the calibration in the used software”?
A: Our calibration report says something like: For the given data, the best calibration model required multiplicative scatter correction for preprocessing, a 2nd derivative Savitsky Golay filtering with 15 points window size, PLS with 5 components, etc.


Data Privacy


Q: Do you offer a package NIR spectra – data based on wet Chemistry?
A: We do not sell NIR-Spectral Data.
We develope calibrations from NIR-Spectral + Lab Data.
See also NIR Calibration Service explained

Q: How can I trust your service does not collect/store/giveaway my data?
A: Please read our Terms of Service (TOS). We do not collect your data.
After processing your data into a calibration, your data is deleted from our servers.
You don’t want to trust anyone? Not even your instrument vendor?
Then anonymize your data, with the NIR JCAMP-Anonymizer Software

Q: How to make others have no re-use of my NIR data?
A: Anonymize your data! Here is how:

Prediction Model Adjustment, Model Correction, Model Updates


Q: Why is it recommended to recalibrate the NIRS on an ongoing basis?
A: read more


Q: How to interpret / judge the Calibration Statistics?
A: Interpretation of the Calibration Statistics