Graph-Based Electrode Selection for Cochlear Implant Channel Interaction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cochlear implants face challenges in programming due to channel interactions, where multiple electrodes stimulate the same neural pathways, leading to poorer hearing outcomes, and current methods for selecting active electrodes are either time-consuming or unreliable, requiring expert review and resulting in suboptimal electrode placement.
Innovation Solution
A graph-based method for active electrode selection in cochlear implants, which estimates activation regions, visualizes overlap, and uses a graph optimization algorithm to identify and deactivate electrodes with significant overlap, ensuring optimal electrode activation and deactivation plans.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple electrodes are activated to provide broader frequency coverage, then the hearing frequency range is improved, but channel interactions occur causing spectral smearing and reduced hearing outcomes
Solution Approach 1:
The patent segments the electrode array into distinct active and inactive groups based on activation region analysis. By dividing the electrodes into functional segments with non-overlapping activation regions, the system achieves both broad frequency coverage through multiple active electrodes and prevents channel interactions by ensuring spatial separation of their neural activation zones.
Solution Approach 2:
The patent applies local quality by customizing the activation status of individual electrodes based on their specific anatomical positions and activation region characteristics. Each electrode is evaluated locally to determine whether it should be active or inactive, allowing the system to optimize frequency coverage in different cochlear regions while preventing overlap in specific local areas where channel interactions would occur.
2Reliability
If manual expert review is used to select active electrodes to avoid channel interactions, then hearing outcomes are improved, but programming time and complexity increase significantly
Solution Approach 1:
The patent implements self-service by enabling the system to automatically perform electrode selection and deactivation decisions without requiring external expert intervention. The computational algorithm autonomously analyzes activation regions, identifies channel interactions, and determines optimal electrode configurations, allowing the cochlear implant system to self-optimize its performance while eliminating time-consuming manual programming processes.
Solution Approach 2:
The patent replaces the mechanical process of manual expert review with a computational algorithm. Instead of relying on human audiologists to visually inspect and manually adjust electrode settings, the system uses automated computational methods to analyze activation regions and make optimization decisions, significantly reducing programming time while maintaining or improving hearing outcomes.
3Reliability
If current state-of-the-art electrode selection methods are used, then some channel interactions are reduced, but only 48% of plans achieve optimality requiring expert review
Solution Approach 1:
The patent implements feedback by using activation region analysis to inform electrode selection decisions. The system calculates the activation region for each electrode based on its anatomical position and electrical characteristics, then uses this feedback information to identify and eliminate channel interactions. This feedback-driven approach ensures that only electrode configurations achieving optimal performance (95%+ success rate) are selected, automatically resolving complex configuration issues without requiring expert intervention.
Data Source
AI summary
A method for active electrode selection in a cochlear implant having an electrode array with a plurality of electrodes implanted in a cochlea of a living subject. The method includes estimating an activation region (AR) of each electrode based on its distance to nerve sites; presenting the AR in a visualization representation, wherein each electrode is represented by a bar having a width or length representing the AR; identifying electrodes having substantial AR overlap if the AR of one electrode overlaps substantially with the AR bar of another electrode; and selecting and deactivating at least one of the identified electrodes with substantial AR overlap.