Arrhythmia Detection With Pre- and Post-Window Noise Counters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current arrhythmia detection algorithms in cardiac monitors suffer from false detections due to noise and low amplitude signals, particularly during periods of high patient activity, leading to suboptimal Positive Predictive Value (PPV) and clinical utility.
Innovation Solution
A method and device for arrhythmia detection that utilizes a grace period before and after the arrhythmia detection window to accumulate and decrement noise events, employing event counters to prevent false detections by considering noise occurrences outside the detection window, thereby enhancing the PPV.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If current noise management algorithms remove or ignore single intervals and counts from arrhythmia detection buffers when a noise event is observed, then the device complexity is reduced, but the Positive Predictive Value (PPV) of arrhythmia detection deteriorates due to false detections during high patient activity or muscle noise
Solution Approach 1:
The algorithm performs preliminary actions by accumulating noise events in a buffer during a predefined window before the arrhythmia detection window. This advance accumulation allows the system to anticipate and prepare for potential noise interference, enabling more accurate differentiation between true arrhythmia events and noise-induced false detections without adding complex real-time processing during the critical detection phase
Solution Approach 2:
The algorithm implements feedback by continuously monitoring noise events during and after the arrhythmia detection window, and using this information to adjust the arrhythmia detection decisions. The noise event accumulation buffer provides feedback about the noise environment, which is then used to validate or reject arrhythmia detections, thereby improving the Positive Predictive Value through iterative refinement based on observed noise patterns
2Reliability
If the algorithm considers noise events occurring after the end of the arrhythmia detection window, then the Positive Predictive Value (PPV) is improved by reducing false positives, but the processing time and computational load increase
Solution Approach 1:
The system performs preliminary accumulation of noise events in a buffer during a predefined window before the arrhythmia detection window. This advance preparation allows the algorithm to have noise event data ready before detection is needed, eliminating the need for complex post-detection analysis and reducing the time penalty for considering post-window noise events
Solution Approach 2:
The algorithm uses periodic action by implementing a structured sequence of predefined windows (pre-detection window, detection window, post-detection window) with specific durations. This periodic structure allows noise accumulation to occur in discrete time blocks, making the extended processing manageable and predictable, thereby reducing overall processing time while maintaining improved PPV
3Reliability
If event counters are used to accumulate and decrement noise events, then false arrhythmia detections are reduced, but the device complexity increases due to additional counting and evaluation mechanisms
Solution Approach 1:
The algorithm segments the noise evaluation process into distinct functional components: a noise event accumulator that counts noise events, a separate arrhythmia detection module, and an evaluation module that combines both inputs. This segmentation allows each component to remain relatively simple while the integrated system achieves high accuracy, as the complexity is distributed across modular, independent units rather than concentrated in a single complex mechanism
Solution Approach 2:
The event counter serves as an intermediary element between the raw noise signal and the arrhythmia detection decision. Rather than directly comparing complex waveforms, the system uses the simplified event counter to mediate the interaction between noise events and detection logic. This intermediary abstraction reduces complexity by transforming continuous analog noise analysis into discrete count-based evaluations that are easier to process
Data Source
Figure 1
Figure 2a
Figure 2b
AI summary
The disclosure relates to a method for arrhythmia detection within a heart signal of a patient, wherein the method is executed by a processor (2) and comprises steps of: providing an input signal (5) which refers to the heart signal of the patient,wherein the input signal (5)comprises cardiac events (6) and noise events (7),wherein the cardiac events (6) are related to cardiac activity of interest of the heart of the patient, and wherein the noise events (7) are not related to cardiac activity of interest of the heart; determining the cardiac events (6) from the input signal (5) in an arrhythmia detection window (9); evaluating the arrhythmia detection window (9) by taking into account at least one of: (i) at least one noise event (7) occurring before a start of the arrhythmia detection window (9), or (ii) at least one noise event (7) occurring after an end of the arrhythmia detection window (9). Also, a device for arrhythmia detection is provided.