Response Standardization for Drift Correction and Multivariate Calibration Transfer in "Electronic Tongue" Studies

Area of Science:

• Chemometrics
• Sensor Technology
• Analytical Chemistry

Background:

• Multivariate calibration for "electronic tongue" (ET) systems requires extensive sample analysis, which is costly and time-consuming.
• Sensor array performance degrades over time due to changes in response characteristics, invalidating predictive models.
• Calibration models are not transferable between different ET systems, even with identical sensors, due to individual sensor variations.

Purpose of the Study:

• To describe procedures for response standardization in "electronic tongue" (ET) studies.
• To present mathematical methods that address sensor drift and enable calibration transfer, avoiding frequent recalibration.
• To reduce the labor and expense associated with maintaining reliable ET predictive models.

Main Methods:

• The study focuses on mathematical methods for sensor response standardization.
• Two popular standardization techniques are described for drift correction.
• These methods facilitate calibration transfer between different "electronic tongue" systems.

Main Results:

• Sensor response standardization effectively corrects for sensor drift over time.
• Calibration transfer between different "electronic tongue" systems is achievable using these mathematical methods.
• The described methods significantly reduce the need for extensive recalibration efforts.

Conclusions:

• Mathematical standardization methods offer a viable alternative to frequent recalibration of "electronic tongue" sensor arrays.
• These techniques address long-term sensor drift and inter-system calibration variability.
• Implementing standardization procedures can lead to more efficient and cost-effective "electronic tongue" data analysis.