Chromatogram Data Processing Using Coarse-to-Fine Dynamic Programming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing chromatogram data processing methods struggle to accurately correct the time axis of chromatograms within a reasonable processing time, especially under conditions with a large number of peaks, high noise, and significant time fluctuations, leading to potential inaccuracies in axis alignment.
Innovation Solution
A chromatogram data processing method utilizing coarse-to-fine dynamic programming, which includes a linear correction step to eliminate linear time deviations, followed by a coarse searching step to narrow down candidates, and a fine searching step to find the optimal correspondence relationship between reference and target chromatograms, effectively reducing the number of calculations and improving precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a dynamic programming algorithm is used to correct nonlinear retention time deviations, then the precision of time axis correction is improved, but the processing time increases exponentially due to the large number of candidate paths
Solution Approach 1:
The patent divides the chromatogram into multiple sections based on peak positions and uses separate dynamic programming calculations for each section. This segmentation reduces the number of candidate paths that need to be evaluated simultaneously, thereby decreasing processing time while maintaining correction precision in each local region
Solution Approach 2:
The patent performs preliminary linear correction of the time axis before applying dynamic programming for nonlinear correction. This preliminary action removes the dominant linear component of retention time deviations, reducing the complexity of the subsequent nonlinear correction and decreasing the number of candidate paths that need to be evaluated
2Reliability
If the beam width in dynamic programming is increased to improve robustness against noise, then the reliability of correction is improved, but the processing time increases due to the larger number of candidates to evaluate
Solution Approach 1:
The patent applies different beam widths to different sections of the chromatogram based on local characteristics such as peak density and noise levels. Sections with high noise tolerance use larger beam widths for robustness, while sections with dense peaks use smaller beam widths for speed, optimizing the balance between reliability and productivity locally
Solution Approach 2:
The patent applies dynamic programming with limited beam width only to critical sections of the chromatogram where nonlinear deviations are most significant, rather than uniformly across the entire chromatogram. This partial application reduces the total number of calculations while maintaining correction quality where it matters most
Data Source
AI summary
Before executing a search for optimal correspondence relationship with a coarse-to-fine DP (dynamic programming) algorithm using time information of peaks appearing in a chromatogram as input data, a simplified linear correction is executed based on detection of start point and end point of the section in which the peaks are present. It is sufficient to correct only nonlinear time deviations in coarse-to-fine DP, and the space to be searched is thus narrowed down. In addition, in coarse stage DP, the number of pieces of data to be processed is reduced by selecting peaks based on peak intensity. Between coarse stage DP and fine stage DP, local inappropriate matching is eliminated by performing filtering processing in accordance with trend of time deviations over the entire chromatogram. The space to be searched in fine stage DP becomes narrow, which reduces the number of candidates to be searched and shortens the calculation time.