Cardiac Electrogram Beat Detection via Multi-Pass Amplitude Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical systems face challenges in accurately analyzing cardiac signal data to distinguish between actual heart beats and noise or inappropriate signals, leading to issues like oversensing and undersensing, which can result in inappropriate therapy delivery.
Innovation Solution
The system employs a multi-pass method for off-line analysis of cardiac electrogram data, using pre-determined amplitude ranges and time intervals to identify and remove candidate beats that are likely noise, thereby improving the accuracy of beat detection and reducing oversensing and undersensing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional single-pass beat detection is used, then the analysis process is simple and fast, but the accuracy of distinguishing actual heart beats from noise is insufficient
Solution Approach 1:
The analysis process is divided into multiple passes, with each pass focusing on specific aspects of beat detection and validation. The first pass identifies candidate beats, the second pass validates them against criteria, and the third pass performs final filtering. This segmentation allows complex accuracy requirements to be met through systematic, staged processing rather than a single complex operation.
Solution Approach 2:
The method performs preliminary beat identification and then systematically validates each candidate against multiple criteria before final acceptance. This preliminary action approach allows the system to process all potential beats thoroughly, ensuring high accuracy by eliminating false positives through predetermined validation rules before the final output is generated.
2Reliability
If aggressive filtering criteria are applied to remove noise, then the purity of detected beats is improved, but the number of false negatives increases
Solution Approach 1:
The filtering criteria are dynamic and adaptive, adjusting based on the specific characteristics of each beat candidate and the patient's baseline data. The system evaluates amplitude, morphology, and temporal relationships flexibly, allowing the filtering intensity to vary rather than applying a fixed aggressive threshold. This dynamic approach maintains high reliability by removing clear false positives while preserving genuine beats that may vary in appearance.
Solution Approach 2:
The multi-pass system incorporates feedback loops where the results of each pass inform the criteria for subsequent passes. Validation rules are refined based on the initial candidate identification, and filtering intensity is adjusted based on the distribution of accepted versus rejected beats. This feedback mechanism ensures the system achieves appropriate therapy delivery criteria while maintaining beat detection completeness through iterative refinement.
Data Source
AI summary
A system and method for performing off-line analysis of cardiac electrogram data, comprising: retrieving an electrogram from a memory location; identifying a first-channel group of candidate beats from at least a first channel of the electrogram; and identifying a second-channel group of candidate beats from at least a second channel of an electrogram. For each first-channel beat candidate near a second-channel beat candidate, the amplitude of the first-channel beat candidate is compared with the amplitude of a previous beat and the amplitude of a next beat on the first electrogram channel, and first-channel beat candidates that are outside of a first pre-determined range from either the previous or next beat are removed. Then first-channel beat candidates that are outside of a second pre-determined range from either the previous or next beat candidate are removed.