Pulse Oximeter Noise Removal via Physiological State-Space Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current biomedical filtering techniques struggle to accurately extract physiological information from signals contaminated by noise and artifacts, especially in mobile and ambulatory patients, due to their reliance on assumptions and approximations about signal and noise spectra, which compromises accuracy and reliability.
Innovation Solution
A probabilistic model using a sigma point Kalman filter or sequential Monte Carlo algorithm combined with Bayesian statistics and dynamic state-space models is employed to remove noise and artifacts from biomedical sensor data, allowing for accurate estimation of physiological parameters like blood oxygen saturation, heart rate, and stroke volume without requiring spectral separation between wanted and unwanted frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional filtering techniques are used, then device complexity is reduced, but measurement precision deteriorates due to inability to accurately separate physiological signals from noise and artifacts
Solution Approach 1:
The patent transforms the filtering problem from frequency-domain operations to state-space parameter estimation. By changing the mathematical representation parameters from spectral frequencies to physiological state variables (heart rate, stroke volume, blood pressure), the system achieves superior signal separation without requiring complex spectral analysis algorithms.
Solution Approach 2:
The patent introduces a physiological model as an intermediary between raw sensor data and extracted parameters. This model acts as a mediator that translates noisy sensor measurements into meaningful physiological parameters through Bayesian inference, avoiding direct complex filtering of the raw signals.
2Reliability
If adaptive filtering based on spectral assumptions is used, then ease of operation is improved, but reliability deteriorates in non-stationary and nonlinear physiological conditions
Solution Approach 1:
The patent employs a dynamic state-space model that adapts to changing physiological conditions in real-time. The model parameters are updated continuously as the patient's physiological state changes, allowing the system to maintain accuracy during exercise, stress, or other non-stationary conditions without manual reconfiguration.
Solution Approach 2:
The Bayesian estimation framework provides continuous feedback between the physiological model and sensor measurements. The system uses measurement data to update the probability distributions of physiological parameters, which in turn refine the filtering of subsequent measurements, creating a self-correcting system that improves reliability over time.
3Loss of information
If simple filters are used, then device complexity is reduced, but loss of information increases because artifacts cannot be reliably distinguished from real physiological processes
Solution Approach 1:
The patent changes the approach from filtering based on frequency characteristics to parameter estimation based on physiological plausibility. By transforming the problem into estimating physiological parameters (heart rate, stroke volume) that have known physiological ranges and relationships, the system preserves genuine signals while rejecting artifacts that violate physiological constraints.
Solution Approach 2:
The patent replaces traditional mechanical filtering approaches (frequency-based filters) with a statistical inference system. Instead of mechanically removing frequency components, the system uses Bayesian probability to infer the most likely physiological parameters from the noisy measurements, preserving information that simple filters would discard.
Data Source
AI summary
A pulse oximeter system comprises a data processor configured to perform a method that combines a sigma point Kalman filter (SPKF) or sequential Monte Carlo (SMC) algorithm with Bayesian statistics and a mathematical model comprising a cardiovascular model and a plethysmography model to remove contaminating noise and artifacts from the pulse oximeter sensor output and measure blood oxygen saturation, heart rate, left-ventricular stroke volume, aortic pressure and systemic pressures.