Wavelet-Based Symbolic Map for ECG Prognostic Indicator Discovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data mining techniques are not designed to discover unrecognized prognostic entities in data, such as ECG data, and lack effectiveness in processing and analyzing time-series data for meaningful insights.
Innovation Solution
A method of signal processing using wavelet transforms to generate a symbolic map of the data, identifying target sequences, and processing the data to obtain waveform prognostic indicators, which involves convolving a wavelet with the data to create a symbolic map and using iterative graphical interfaces like MATLAB for analysis and optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing data mining techniques are used, then data processing capability is improved, but the ability to discover unrecognized prognostic entities remains insufficient
Solution Approach 1:
The patent segments the data processing into distinct stages: wavelet transform to decompose the signal into frequency components, symbolic map generation to represent patterns, target sequence identification to find prognostic indicators, and contextual processing to interpret findings. This segmentation enables both efficient processing and discovery of unrecognized patterns by treating each stage independently optimizable
Solution Approach 2:
The patent introduces a symbolic map as an intermediary representation between raw data and prognostic indicators. The symbolic map translates complex waveforms into recognizable patterns and sequences, serving as a mediator that enables both processing efficiency and pattern discovery. This intermediary layer allows the system to handle complex data while identifying prognostic entities that would be difficult to detect directly
2Measurement precision
If wavelet transforms are applied to generate symbolic maps, then pattern recognition is improved, but computational complexity increases
Solution Approach 1:
The patent applies wavelet transforms with different scales and frequencies to different portions of the data, focusing computational resources where needed. By using local quality analysis, the system achieves high pattern recognition accuracy in critical regions while reducing overall computational complexity through selective processing
Solution Approach 2:
The patent performs preliminary wavelet transform and symbolic map generation before the actual prognostic indicator identification. This preliminary action pre-processes the data into a format that simplifies subsequent analysis, reducing the computational complexity of the main processing task while maintaining high pattern recognition accuracy
3Adaptability or versatility
If iterative graphical interfaces are used for analysis, then optimization capability is improved, but processing time increases
Solution Approach 1:
The patent implements iterative graphical interfaces that operate in periodic cycles, where each iteration refines the analysis based on previous results. This periodic action allows the system to achieve high optimization capability by progressively improving pattern recognition while managing processing time through structured iteration rather than continuous computation
Data Source
AI summary
A method of signal processing of data using a computer by generating a symbolic map of the data using at least one wavelet transform, identifying target sequences in the symbolic map; and processing the data with reference to the target sequences to obtain waveform prognostic indicators.