Probabilistic Digital Signal Processor for Biomedical Sensor Noise Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mechanical and biomedical sensors face challenges in filtering out contaminating signals, particularly those caused by patient movement and nonlinear, non-stationary noise, which complicates the extraction of reliable physiological information, and they typically measure a limited range of medical parameters.
Innovation Solution
A probabilistic model-based system that uses fused data from multiple sensors to filter, estimate, and extract additional information by integrating a dynamic state-space model with a probabilistic digital signal processor, enabling the processing of both mechanical and biomedical data to provide clearer and more comprehensive insights.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional filtering techniques are used to remove noise and artifacts from sensor data, then some noise reduction is achieved, but reliable physiological information cannot be extracted when artifacts resemble real processes (e.g., ectopic beats)
Solution Approach 1:
The patent introduces a probabilistic model as an intermediary layer between the raw sensor data and the final physiological parameter extraction. This model incorporates prior knowledge about physiological processes and noise characteristics to mediate the interpretation of ambiguous signals, allowing the system to distinguish between artifacts and real physiological events even when they resemble each other
Solution Approach 2:
The patent transforms the filtering problem from a deterministic signal processing task into a probabilistic parameter estimation problem. By changing the approach from fixed filtering thresholds to adaptive probabilistic parameters that evolve with the data, the system can reliably extract physiological information even when artifacts mimic real processes
2Productivity
If mechanical and biomedical sensors are used to monitor physiological parameters, then real-time monitoring is achieved, but the sensors are limited to measuring only a narrow spectrum of medical parameters
Solution Approach 1:
The patent creates a universal probabilistic framework that can process and extract multiple types of physiological parameters from the same sensor data stream. The system is designed to be multi-functional, capable of monitoring various biomedical states simultaneously by adapting the probabilistic model to different physiological parameters of interest
Solution Approach 2:
The patent implements a dynamic parameter estimation approach where the system can adaptively adjust which parameters are being monitored and how they are extracted in real-time. The probabilistic model dynamically updates its estimates based on incoming data, allowing the system to expand its measurement capabilities beyond a fixed set of parameters
3Reliability
If sensors operate in ambulatory and physically active patients to capture real-world data, then ecological validity is improved, but the data becomes corrupted by severe artifacts such as signal dropouts and large periodic artifacts
Solution Approach 1:
The patent converts the harmful motion artifacts into useful information by using them as inputs to the probabilistic model. Rather than attempting to completely eliminate motion artifacts, the system learns to interpret them in context, using the probabilistic framework to distinguish between motion-related signal changes and actual physiological events, thereby maintaining ecological validity while managing artifact corruption
Data Source
AI summary
A probabilistic digital signal processor using data from multiple instruments is described. In one example, a digital signal processor is integrated into a biomedical device. The processor is configured to: use a dynamic state-space model configured with a physiological model of a body system to provide a prior probability distribution function; receive sensor data input from at least two data sources; and iteratively use a probabilistic updater to integrate the sensor data as a fused data set and generate a posterior probability distribution function using all of: (1) the fused data set; (2) an application of Bayesian probability; and (3) the prior probability distribution function. The processor further generates an output of a biomedical state using the posterior probability function.