Multichannel Heartbeat Detection via Temporal Pattern Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cardiac monitoring technologies face challenges in accurately detecting heartbeats in noisy multi-channel signals, particularly when noise peaks resemble true beats, leading to high false positive or false negative rates, and existing methods for combining multiple channels are suboptimal.
Innovation Solution
The approach involves combinatorial optimization of peak sequences across multiple channels, scoring them based on quality measures such as peak timing coherence and prominence, and selecting high-quality sequences to improve heartbeat detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional peak detection with simple thresholding is used in noisy ECG signals, then detection speed is maintained, but false positives and false negatives increase significantly when noise peaks resemble true beats
Solution Approach 1:
The patent segments the heartbeat detection problem into multiple independent quality measures (peak prominence, timing coherence, morphological similarity, signal-to-noise ratio) that are computed separately and then combined. This segmentation allows each measure to be optimized independently while maintaining overall detection accuracy in noisy conditions.
Solution Approach 2:
The patent changes the detection parameters by introducing multiple quality metrics beyond simple thresholding. Instead of a single threshold parameter, the system uses peak prominence, timing coherence, morphological similarity, and signal-to-noise ratio, allowing adaptive adjustment of detection criteria based on signal characteristics.
2Measurement precision
If simple threshold-based fusion schemes are used across multiple channels, then computational simplicity is maintained, but detection accuracy deteriorates when ECG detection criteria are too broad or narrow
Solution Approach 1:
The patent implements feedback mechanisms where detection results from one channel inform the detection process in other channels. The quality measures computed from multiple channels are used to adjust detection thresholds and criteria dynamically, creating a feedback loop that improves precision while managing complexity through iterative refinement.
Solution Approach 2:
The patent combines multiple detection approaches and quality measures into a composite detection framework. By fusing peak prominence, timing coherence, morphological similarity, and signal-to-noise ratio measures across multiple channels, the system creates a composite detection mechanism that achieves higher precision than any single method alone.
3Reliability
If autocorrelation Bayesian probability is used to fuse multiple channels, then probabilistic fusion is achieved, but performance deteriorates in high noise conditions where noise peaks and true beats have similar sizes and shapes
Solution Approach 1:
The patent applies local quality assessment by computing quality measures for individual peaks rather than relying solely on global autocorrelation probabilities. Each peak is evaluated for prominence, timing coherence, and morphological characteristics, allowing discrimination of local features that distinguish true beats from noise peaks even when their overall sizes are similar.
Solution Approach 2:
The patent adds additional dimensions to the fusion process beyond simple probability multiplication. By incorporating temporal dimension (timing coherence across channels) and morphological dimension (waveform similarity), the system creates a multi-dimensional discrimination space that separates true beats from noise peaks more effectively than single-dimensional autocorrelation alone.
Data Source
AI summary
A method for detecting heart beats within multichannel cardiac signals is disclosed. A plurality of sensors are configured to receive multiple channel cardiac signals. A processor is configured to preprocess the cardiac signal and then detect heart beats by separately analyzing signal segments. Separately in each channel, candidate peaks are selected. Peaks across channels that are closely time aligned are merged, and a peak timing coherence probability is computed for each merged peak. A sequence search is then performed on global peaks, which comprise the merged peaks and the remainder of the candidate peaks (which exist only in a single channel). Resulting sequences are assigned raw scores based on: 1) the sum of the sequence's peak coherence probabilities; 2) the sum of peak pair prominence scores across all channels; 3) rhythm probability, which is temporal regularity in the case of sinus rhythm; and 4) the number of skipped beats.