Interactive ECG Signal Processing with Selectable Digital Filters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ECG over-reading software lacks the ability to apply arbitrary noise filters and compare their results, leading to potential incorrect diagnoses due to noise artifacts, especially when recording the P-wave or atrial signal.
Innovation Solution
A system and method for interactive processing of ECG data that allows users to select and apply multiple digital filters to specific portions of an ECG waveform, display filtered results side-by-side for comparison, and recommend filters based on noise frequency, enabling effective noise reduction and improved diagnostic accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional ECG over-reading software uses proprietary filters only, then the software is simple to operate, but the user cannot apply arbitrary noise filters of choice to remove specific noise artifacts
Solution Approach 1:
The software divides the filtering process into separate selectable filter modules (e.g., power line noise filter, baseline wander filter, muscle noise filter). Each filter can be independently selected and applied to specific portions of the ECG trace, allowing users to customize the filtering approach without overwhelming complexity.
Solution Approach 2:
The software provides dynamic filter selection where users can choose different filters based on the specific noise characteristics observed in the ECG recording. The system adapts to different recording conditions by allowing real-time selection of appropriate filters rather than using a fixed proprietary filter set.
2Measurement precision
If conventional ECG software provides only a single filtered output, then the processing is simple, but the user cannot compare results of different noise filters to find the most appropriate one
Solution Approach 1:
The software creates multiple copies of the ECG trace, each processed with a different filter selection. These copied and filtered traces are displayed side-by-side or in a comparative view, allowing users to visually compare the effectiveness of different filters on the same underlying data without requiring complex reprocessing.
Solution Approach 2:
The system applies multiple filters beyond what a single clinical scenario might require, allowing users to select the most appropriate one. This excessive filtering capability ensures that the correct filter is available even if the initial choice is not optimal, and comparison makes it easy to identify the best fit.
3Object-affected harmful factors
If noise filters are applied to entire ECG recordings, then noise reduction is comprehensive, but important diagnostic information may be lost due to over-filtering
Solution Approach 1:
The software enables application of different filters to different portions or segments of the ECG recording. Users can identify specific segments containing noise artifacts and apply targeted filtering only to those portions, while leaving other segments unfiltered or minimally filtered to preserve diagnostic information.
Solution Approach 2:
The ECG recording is divided into manageable segments or portions that can be independently filtered. This segmentation allows selective application of filtering strength - aggressive filtering where noise is present, and minimal or no filtering where diagnostic signals are critical.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A system (40) and method (60) for interactive processing of ECG data (10) are presented. An electrocardiogram is displayed (63). A user selection of a portion of the displayed ECG (10) is received (64). Digitized signals corresponding to the selection are obtained (65). A list of digital filters (56) for filtering the selection are displayed (67). A user selection of one or more sets of the digital filters (56) is received (69), with each of the sets including one or more of the filters (56) from the list. The selected sets are applied (70) to the digitized signals for the selection. A filtered ECG (10) for the selection is generated (71) for each of the sets based on the signals filtered by that set. The filtered selection ECG for each of the sets are presented (71) on the display.