Iterative Spectral Matching for Multicomponent Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Identifying unknown spectra from spectrometer measurements is time-consuming due to differences in measurement conditions and the need to compare with large reference libraries, often requiring hours or days, even with high-speed processors, especially when dealing with mixtures that don't match single reference spectra but rather combinations of them.

Innovation Solution

A method that iteratively compares an unknown spectrum to candidate spectra, generating new combinations of reference spectra to find matches, reducing computational load by pruning less similar spectra and presenting ranked candidate spectra with their corresponding weights, allowing for faster and more accurate identification.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If exhaustive comparison of unknown spectrum to all possible combinations of reference spectra is performed, then identification accuracy is improved, but computational time increases significantly

Engineering Contradiction:
Improveidentification accuracyVSAvoidcomputational time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the exhaustive search space by dividing reference spectra into groups and performing iterative matching. Instead of comparing unknown spectra to all possible combinations simultaneously, the method breaks down the problem into smaller sub-problems by processing spectra in manageable groups, thereby reducing computational time while maintaining identification accuracy.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies preliminary action by performing preprocessing steps on reference spectra before the actual matching process. Reference spectra are pre-processed, pre-grouped, and organized into structures that facilitate faster comparison during the matching phase, reducing the computational burden during the actual identification process.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If quantitative analysis with regression is performed to determine weighting of each reference spectrum, then analysis precision is improved, but computational complexity increases

Engineering Contradiction:
Improvequantitative analysis precisionVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the regression computation by performing it iteratively on smaller subsets of reference spectra rather than all spectra simultaneously. This divides the complex regression problem into smaller, more manageable computational tasks that can be processed sequentially, reducing overall computational complexity while maintaining precision.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies partial action by performing regression analysis on a selected subset of reference spectra that are most relevant to the unknown spectrum, rather than performing regression on all reference spectra. This partial processing reduces computational complexity while still achieving the required analysis precision for quantitative determination.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP2245443B1Efficient spectral matching, particularly for multicomponent spectra
Publication Date: 2018.12.12 THERMO ELECTRONICS SCI INSTR LLC
  • EP2245443B1 patent drawingFigure 1
  • EP2245443B1 patent drawingFigure 2
  • EP2245443B1 patent drawingFigure 3

AI summary

An unknown spectrum obtained from infrared or other spectroscopy can be compared to spectra in a reference library to find the best matches. The best match spectra can then each in turn be combined with the reference spectra, with the combinations also being screened for best matches versus the unknown spectrum. These resulting best matches can then also undergo the foregoing combination and comparison steps. The process can repeat in this manner until an appropriate stopping point is reached, for example, when a desired number of best matches are identified, when some predetermined number of iterations has been performed, etc. This methodology is able to return best-match spectra (and combinations of spectra) with far fewer computational steps and greater speed than if all possible combinations of reference spectra are considered.