Form Parameter Forecaster for Overlapping Cardiac Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Implantable medical devices face challenges in accurately processing cardiac signals distorted by noise, such as Far Field R-waves, which can overlap with P-waves, leading to distorted waveforms that existing digital signal processing techniques struggle to distinguish.
Innovation Solution
The use of form parameter forecasting, where composite waveforms are generated by superimposing P-wave and Far Field R-wave templates with different time shifts, allowing for the derivation of form parameters and creation of a multidimensional map to identify the signal component of interest, even in overlapping scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If digital signal processing techniques use form factor histograms to classify sensed atrial signals, then classification between P-wave and Far Field R-wave is improved, but the technique fails when waveforms overlap or near overlap causing distorted signals
Solution Approach 1:
The patent pre-calculates and stores form parameter values for composite waveforms representing all possible combinations of P-waves and Far Field R-waves at different time shifts. This preliminary action creates a comprehensive reference library that enables accurate classification even when waveforms overlap, eliminating the need for real-time complex calculations.
Solution Approach 2:
The patent extends the analysis from traditional single-dimension form factor histograms to a multidimensional space by incorporating multiple form parameters (e.g., area, width, height, slope) and comparing them against a library of composite waveforms with varying time shifts. This dimensional expansion allows the system to distinguish overlapping waveforms that cannot be separated by traditional methods.
2Productivity
If P-wave and Far Field R-wave occur close in time, then the resulting waveform is a combination of both components, but the waveform becomes distorted and does not resemble either component
Solution Approach 1:
The patent creates copies of template waveforms (P-waves and Far Field R-waves) at various time shifts and combines them to generate composite waveform templates. These copied and combined templates serve as reference patterns against which actual sensed signals are compared, enabling accurate identification even when the original signal components are distorted by overlap.
Solution Approach 2:
The patent systematically varies the time shift parameter between P-wave and Far Field R-wave templates to generate a library of composite waveforms. By changing this temporal parameter across multiple discrete values, the system creates a comprehensive set of reference patterns that cover all possible overlap scenarios, allowing accurate matching regardless of the actual time relationship between components.
3Reliability
If traditional signal processing requires time separation between P-wave and R-wave histograms, then each histogram has a specific form, but this requirement limits applicability when separation is not present
Solution Approach 1:
The patent creates a universal classification system that handles both separated and overlapping waveforms through a single methodology. The form parameter forecaster and composite waveform library approach works regardless of whether P-waves and Far Field R-waves are separated in time or overlapping, making the system universally applicable to all signal conditions without requiring different processing paths.
Solution Approach 2:
The patent transitions from static histogram analysis (which assumes fixed, separated waveforms) to a dynamic approach that accounts for variable time relationships between P-waves and Far Field R-waves. By incorporating time-shifted composite waveforms into the reference library, the system adapts to dynamic signal conditions where the temporal relationship between components may vary from beat to beat.
Data Source
AI summary
Waveform analysis is used to identify and distinguish components of a sensed input signal, such as P-wave and Far Field R-wave signal components present in a sensed cardiac signal, even when the components are so closely spaced in time that the overlap to create a distorted input signal. A set of composite waveforms are generated by superimposing waveform templates of the signal components with different time delays or degree of overlap. Form parameters for each composite waveform are derived and mapped in a multidimensional map, from which form parameter boundaries are derived. Waveform data is collected from an input signal during a sensed event time window, and form parameters for the input signal waveform are derived. An output identifying the signal component of interest (e.g., a P-wave) and its location within the sensed event time window is produced based upon the set of form parameters of the input signal waveform and the form parameter boundaries.