ECG Artifact Detection via Signal Segmentation and Baseline Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ECG signal analysis systems face challenges in accurately detecting and rejecting artifact signals, leading to false arrhythmia alarms and reduced precision in heart rate calculation and beat classification due to noise contamination.
Innovation Solution
An apparatus with processors configured to select sample points, extract features, and apply transformations to detect artifact signals in real-time, preventing false identification of artifacts as part of the QRS-complex and improving the accuracy of heart rate calculation and alarm generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If artifact detection and rejection algorithms are implemented in ECG signal analysis, then measurement precision and reliability of cardiac event detection are improved, but device complexity and computational requirements increase
Solution Approach 1:
The artifact detection system segments the ECG signal into multiple components (QRS complex, P-wave, T-wave, baseline) and analyzes each segment separately using specific algorithms tailored to each component's characteristics. This segmentation allows precise artifact detection in each segment without requiring a single complex algorithm to handle all signal types.
Solution Approach 2:
The patent introduces an intermediary baseline correction step that removes low-frequency drift and artifacts before subsequent analysis. This intermediary processing stage simplifies the overall system by preprocessing the signal, making later detection algorithms more effective and reducing their complexity requirements.
2Reliability
If real-time artifact detection is performed on all ECG signals, then reliability of cardiac monitoring is improved, but processing time and loss of time increase
Solution Approach 1:
The system performs preliminary artifact detection and baseline correction on incoming ECG signals before full analysis. By preprocessing signals to remove obvious artifacts and correct baseline drift early in the pipeline, the system reduces the computational burden on subsequent real-time analysis stages, maintaining reliability while reducing processing time.
Solution Approach 2:
The patent applies different levels of artifact detection scrutiny to different signal segments based on their importance. Critical segments like the QRS complex receive more intensive analysis while less critical segments undergo lighter processing. This partial action approach maintains reliability for critical detections while reducing overall processing time.
3Measurement precision
If multiple artifact detection algorithms are applied to differentiate physiological and non-physiological artifacts, then measurement precision is improved, but device complexity and difficulty of detecting and measuring increase
Solution Approach 1:
The patent applies different detection algorithms and criteria to different types of artifacts based on their local characteristics. Physiological artifacts (muscle tremor, respiration) are detected using algorithms sensitive to specific frequency ranges and morphological patterns, while non-physiological artifacts (electromagnetic interference, baseline drift) are detected using different criteria. This local quality approach enables precise artifact classification without requiring a single overly complex universal detector.
Data Source
AI summary
A method of processing of electrocardiogram (“ECG”) signals from at least one ECG lead connected to a patient includes computing an average beat from a plurality of beats occurring during a predetermined averaging interval. An onset point, isoelectric point, and a J-point of the average beat is computed to establish an S-T segment of the average beat. The isoelectric point and the J-point of the average beat are used to determine an S-T segment for each beat of the averaging interval, and the average of the S-T segments of each beat is computed and averaged with the S-T segment of the average beat.