Automated Calibration Curve Generation in Mass Spectrometry

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for generating calibration curves in mass spectrometry are labor-intensive and prone to errors due to the need for manual exclusion of outliers, especially when dealing with large datasets or samples where the number of outliers exceeds the number of valid points, leading to improper models that are not representative of the analytical value to concentration relationship.

Innovation Solution

An automated method that selects calibration point pairs at consecutive concentration levels, performs linear regression, and iteratively adds points that improve the R² value, ensuring a subset of points that best fits a predefined curve fit type, such as linear or quadratic, while excluding outliers based on accuracy and residual metrics.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual exclusion of outliers is performed to generate accurate calibration curves, then measurement precision is improved, but productivity deteriorates due to labor-intensive processing

Engineering Contradiction:
Improvecalibration curve accuracyVSAvoidprocessing speed
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system automatically identifies and excludes outlier points from calibration curves without requiring manual intervention. The software performs statistical analysis, calculates residuals, and removes points that deviate significantly from the fitted model, enabling the system to self-correct and generate accurate calibration curves autonomously.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The manual mechanical process of visually inspecting and removing outlier points is replaced with an automated computational system. The software uses algorithms to calculate residuals, apply statistical tests (such as the 3-sigma rule or studentized residuals), and automatically exclude outliers, substituting human judgment with systematic computational methods.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If automated outlier removal methods are applied to large datasets, then productivity is improved, but measurement precision deteriorates when outliers exceed valid points

Engineering Contradiction:
Improveprocessing speedVSAvoidcalibration curve accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The system employs iterative feedback mechanisms where initial calibration curves are generated, residuals are calculated, and outlier points are identified. The system then removes these outliers and regenerates the calibration curve, using the new curve to identify additional outliers, repeating this process until convergence is achieved. This feedback loop ensures that even when outliers exceed valid points, the final model accurately represents the true relationship.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The outlier detection and removal process is dynamic rather than static. The system adapts its criteria for outlier identification based on the distribution of residuals and the specific characteristics of each dataset. It can adjust sensitivity thresholds and apply different statistical tests depending on the proportion of outliers detected, allowing the method to remain effective even when outliers outnumber valid points.

Inventive Principle:
Principle #15Dynamics

3Productivity

If all data points are used to generate calibration curves, then productivity is improved, but measurement precision deteriorates due to inclusion of outliers

Engineering Contradiction:
Improveprocessing speedVSAvoidcalibration curve accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The system extracts and removes outlier points from the dataset before generating the final calibration curve. By identifying points with large residuals or those that fail statistical outlier tests, the system separates these problematic points from the valid data, ensuring that only representative points contribute to the final model while maintaining efficient automated processing.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP3250917B1Automatic quantitative regression in a mass spectrometer
Publication Date: 2022.01.12 DH TECH DEVMENT PTE
  • EP3250917B1 patent drawingFigure 1~2
  • EP3250917B1 patent drawingFigure 3~4
  • EP3250917B1 patent drawingFigure 5~6

AI summary

A method of creating a calibration curve. The method involves analyzing calibration standards and obtaining analytical values at a concentration level and then determining an initial series of consecutive points that best fits a defined calibration model. The series of points are then grown by considering the next available point at the next available concentration level, either higher or lower and considering whether the inclusion of such a point would be suitable. The process is then repeated until all points have been considered.