Cochlear Implant Channel Selection Temporal Modification

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional N-of-M channel selection strategies for cochlear implant systems either lose information in lower amplitude channels or are computationally intensive, failing to provide complete spectral coverage and efficiently utilize temporal masking phenomena.

Innovation Solution

Implementing a temporal modification approach in the sound processor to increase the probability of selecting channels not previously selected and decrease the probability of selecting channels that were recently active, ensuring complete spectral coverage while being computationally efficient.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If N-of-M channel selection strategies select only the analysis channels with the highest amplitude signals for presentation to a patient during a particular stimulation frame, then the system can present the most prominent signals to the patient, but all information in the lower amplitude channels is lost during that frame

Engineering Contradiction:
Improvesignal amplitude detection accuracyVSAvoidinformation in lower amplitude channels
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent applies periodic action by using temporal masking approaches that vary channel selection probability over time based on recent selection history. Channels that were not selected in previous stimulation frames have their selection probability increased, creating a periodic opportunity for previously muted channels to be presented, thus preventing permanent information loss while maintaining amplitude-based selection when channels are relevant.

Inventive Principle:
Principle #19Periodic action

2Reliability

If temporal masking approaches are used to select the most relevant channels for presentation to a patient during a particular stimulation frame, then the system takes into account the refractory phenomena associated with auditory nerve fibers, but these approaches are computationally intensive, requiring computation of masking models during each stimulation frame

Engineering Contradiction:
Improvetemporal masking accuracyVSAvoidcomputational complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent changes the parameter of channel selection from deterministic amplitude-based selection to probability-based selection with temporal modification. By modifying selection probabilities based on recent selection history rather than computing full masking models, the system maintains temporal masking accuracy while significantly reducing computational complexity to simple probability updates.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If temporal masking approaches decrease the probability that a particular channel that has been selected for presentation during a stimulation frame will be again selected for presentation during one or more stimulation frames that immediately follow, then the system accounts for auditory nerve fiber refractory phenomena, but these approaches do not specifically increase the probability that a particular channel that has not been selected for a relatively long time will be again selected

Engineering Contradiction:
Improvetemporal masking effectivenessVSAvoidinformation in channels not selected for extended periods
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent implements feedback by monitoring which channels have been selected or not selected over recent stimulation frames and using this history to dynamically adjust selection probabilities. Channels not selected for extended periods receive increased selection probability, creating a feedback loop that ensures comprehensive spectral coverage while respecting temporal masking constraints.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3204108B1Channel selection systems that employ temporal modification
Publication Date: 2018.12.05 ADVANCED BIONICS AG
  • EP3204108B1 patent drawingFigure 1
  • EP3204108B1 patent drawingFigure 2
  • EP3204108B1 patent drawingFigure 3

AI summary

An exemplary sound processor may 1) divide an audio signal into M analysis channels, 2) select only N analysis channels included in the M analysis channels for presentation to a patient during a stimulation frame, wherein N is less than M, 3) increase a probability that a particular analysis channel will be included in the N analysis channels selected for presentation to the patient during the stimulation frame if the particular analysis channel was not selected for presentation to the patient during one or more stimulation frames that temporally precede the stimulation frame, and 4) decrease the probability that the particular analysis channel will be included in the N analysis channels selected for presentation to the patient during the stimulation frame if the particular analysis channel was selected for presentation to the patient during the one or more stimulation frames that temporally precede the stimulation frame.