Probabilistic Entropy for Neurostimulation Artifact Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for electrical stimulation therapy in medical devices face challenges in accurately identifying and filtering out artifacts from bioelectrical signals, which can lead to erroneous measurements and reduced effectiveness of treatments like adaptive deep brain stimulation (aDBS).
Innovation Solution
The use of probabilistic entropy analysis to differentiate between clean bioelectrical signals and those contaminated with artifacts, allowing for the selection of electrodes that provide high-quality signals for therapy control, by determining the entropy values of sensed signals and comparing them to established thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional algorithms are used to detect artifacts, then the device complexity is reduced, but the measurement precision deteriorates due to inability to detect periodic artifacts with variable amplitudes
Solution Approach 1:
The patent transforms the artifact detection problem by changing the analysis parameter from amplitude-based detection to entropy-based detection. By computing probabilistic entropy of signal features, the system can detect periodic artifacts regardless of their amplitude variations, resolving the contradiction between detection accuracy and algorithm complexity.
Solution Approach 2:
The patent replaces conventional signal processing mechanisms with information-theoretic approaches. Instead of using traditional filters or threshold-based detectors, the system uses entropy calculation to identify periodic patterns, achieving superior detection capability with moderate computational complexity.
2Reliability
If all electrodes are used for therapy delivery, then the productivity is increased, but the reliability deteriorates due to inclusion of electrodes with artifact-contaminated signals
Solution Approach 1:
The patent segments the electrode array into two subsets: those with clean signals (entropy above threshold) and those with artifact-contaminated signals (entropy below threshold). This segmentation allows the system to maintain high reliability by excluding compromised electrodes while preserving productivity by utilizing valid electrodes for therapy delivery.
Solution Approach 2:
The system implements feedback by continuously monitoring signal entropy and using this information to dynamically select which electrodes to use for therapy. The entropy measurement provides real-time feedback on signal quality, enabling adaptive electrode selection that balances reliability and productivity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Techniques are disclosed for using probabilistic entropy to select electrodes with fewer artifacts for controlling adaptive electrical neurostimulation. In one example, a plurality of electrodes sense bioelectrical signals of a brain of a patient. Processing circuitry determines, for each bioelectrical signal sensed at a respective electrode of the plurality of electrodes, a probabilistic entropy value of the bioelectrical signal. The processing circuitry compares each of the respective probabilistic entropy values of the bioelectrical signal to respective entropy threshold values and selects, based on the comparisons, a subset of electrodes of the plurality of electrodes. The processing circuitry controls, based on the bioelectrical signals sensed via respective electrodes of the subset of electrodes and excluding the bioelectrical signals of the plurality of bioelectrical signals sensed via respective electrodes not in the subset of electrodes, delivery of electrical stimulation therapy to the patient.