Ion Mobility Spectrometry Library Segmentation for Substance Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ion mobility spectrometry methods face challenges in reliably identifying substances, particularly explosives, due to interference from other substances, leading to false alarms and limited accuracy caused by diffusion broadening and sensor instability, which complicates the comparison of mobility spectra.
Innovation Solution
The method involves acquiring and comparing series of ion mobility spectra, dividing the reference library into classes based on easily computable characteristics like spectrum moments, and using idealized spectra for similarity comparisons to reduce computational effort and increase identification certainty.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If complete series of spectra are acquired and compared for identification, then identification certainty is improved, but computational effort and processing complexity increase significantly
Solution Approach 1:
The reference library is divided into multiple classes based on easily computable characteristics such as spectrum moments (e.g., drift time, peak area). This segmentation allows the system to first classify the sample spectrum into a specific class and then only compare it with reference spectra from that class, dramatically reducing the number of comparisons needed while maintaining high identification certainty
Solution Approach 2:
Different parts of the spectral data are used for different purposes: easily computable characteristics (spectrum moments) are used for classification, while the complete spectral information is used for final identification within the classified group. This local quality approach optimizes both speed and accuracy
2Measurement precision
If mobility resolution is improved to achieve better identification, then measurement precision is improved, but diffusion broadening limits the achievable resolution
Solution Approach 1:
The system transforms the spectral data into different parameter representations, specifically using spectrum moments (drift time, peak area) as key parameters. This parameter transformation allows the system to work with the information that is least affected by diffusion broadening while still capturing the essential identification features
Solution Approach 2:
Instead of directly comparing complete spectral profiles which are affected by diffusion broadening, the system creates simplified representations (spectrum moments) that capture the essential characteristics. These moment-based representations serve as robust copies for comparison that are less sensitive to diffusion effects
3Measurement precision
If tolerance for mobility comparison is reduced to improve accuracy, then measurement precision is improved, but sensor instability prevents achieving such tight tolerances
Solution Approach 1:
The system changes from using absolute mobility values (which are sensitive to sensor instability) to using spectrum moments and their ratios. These transformed parameters are relative measures that are much less sensitive to drift and instability in the sensor system, allowing for more reliable comparisons without requiring tight tolerances on absolute values
4Reliability
If more peaks are used for identification to improve reliability, then identification certainty is improved, but the number of detectable peaks is limited by mobility resolution
Solution Approach 1:
The system moves from one-dimensional peak comparison (drift time only) to a multi-dimensional approach using spectrum moments (drift time, peak area, and their combinations). This dimensional expansion allows the system to extract more identification information from the same spectral data without requiring higher mobility resolution
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the reliability of substance identification by reducing false alarms and improving accuracy, allowing for efficient identification of substances even in complex mixtures, while maintaining manageable computational processing.
Implementation Method 1
The ions of the substances are usually formed by so-called 'chemical ionization at atmospheric pressure' (APCI) in reactions with reactant ions, usually by protonation or deprotonation of the substance molecules or by electron transfer or ion attachment
Implementation Method 2
An axial electric field pulls the ions of these ion pulses through a stationary drift gas in the drift region, their velocity being determined by their 'mobility'
Implementation Method 3
At the end of the drift region, the incident ion current is measured at an ion detector, digitized and stored as a 'mobility spectrum'
Data Source
AI summary
The invention relates to a method and an apparatus for identifying substances in a sample by similarity comparisons between series of ion mobility spectra of the sample and series of ion mobility spectra of reference samples. The invention comprises dividing the collection of series of spectra of reference samples - termed reference library - into classes, with the class assignment of a series of spectra being calculated from the measured values themselves, and limiting the similarity comparisons to series of spectra with the same class assignment. First and second moments of the spectra have proven to be particularly favorable characteristics for the class assignment. In a preferred embodiment, only mobility spectra of the series of spectra with the same first moments are compared with each other in the similarity comparisons.