Binaural Cochlear Implant Sound Object Segregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cochlear implant users face difficulties in auditory scene analysis, particularly in isolating and processing individual sound sources in noisy environments, leading to reduced speech intelligibility and degraded musical perception due to the inability to effectively segregate sound objects.
Innovation Solution
A bilateral cochlear implant system with sound processing arrangements that include left-side and right-side sensing microphones, a sound object identification module, and a stimulation side selector module to analyze and process sound objects, adjusting phase and amplitude components to selectively stimulate the appropriate ear or ears for improved sound localization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional cochlear implant systems process all audio signals equally through a single channel, then the system structure remains simple, but the ability to segregate sound objects and achieve auditory scene analysis deteriorates
Solution Approach 1:
The patent divides the audio signal processing into multiple frequency channels, with each channel independently analyzing and processing specific sound objects. This segmentation allows the system to separate different sound sources (speech, music, noise) across different frequency bands, improving sound object segregation while managing complexity through structured channel-based processing
Solution Approach 2:
Different processing strategies are applied to different frequency channels based on their characteristics. The system applies monaural cues (harmonicity, common onset) and binaural cues (ITD, ILD) selectively to specific channels, allowing localized optimization of sound object identification without requiring complete reprocessing of all signals
2Measurement precision
If the system uses only monaural cues for sound object identification, then the processing complexity remains lower, but the accuracy of sound localization and auditory scene analysis deteriorates
Solution Approach 1:
The patent combines multiple types of cues (monaural and binaural) within each frequency channel to identify sound objects. By merging harmonicity analysis, common onset detection, ITD measurement, and ILD measurement, the system achieves superior sound object identification accuracy that leverages the complementary strengths of different cue types
Solution Approach 2:
The system dynamically selects and weights different cues based on the specific acoustic environment and signal characteristics. When binaural cues are more reliable (e.g., for localization), the system emphasizes ITD and ILD; when monaural cues are more useful (e.g., for speech identification), the system prioritizes harmonicity and onset analysis
3Loss of information
If the system processes sound objects in all frequency channels with equal detail, then the completeness of auditory scene analysis improves, but the computational load and processing time increase
Solution Approach 1:
The system applies full sound object identification processing only to frequency channels where it is most beneficial (typically mid-frequency ranges containing speech and important auditory information). In less critical frequency ranges, the system applies simplified processing or relies on information from adjacent channels, reducing overall computational load while maintaining adequate coverage
Solution Approach 2:
The system performs preliminary analysis of the acoustic environment to identify dominant sound objects and their frequency distributions before detailed processing. This preliminary step allows the system to pre-configure which channels require intensive processing and which can use streamlined algorithms, optimizing processing time based on actual scene requirements
4Measurement precision
If the system provides separate processing for each ear independently, then the simplicity of unilateral processing is maintained, but the binaural advantages for sound localization and auditory scene analysis are lost
Solution Approach 1:
The system uses feedback from interaural time differences and interaural level differences to continuously adjust and refine sound object identification and localization. By comparing the timing and level of signals arriving at each ear, the system generates feedback that improves the accuracy of sound object segregation and spatial positioning, with each ear's processing informing the other
Data Source
AI summary
A sound processing arrangement is described for a patient with a bilateral cochlear implant system having implanted electrode arrays in each ear. There is a left-side sensing microphone and a right-side sensing microphone, each configured for sensing the sound environment surrounding the patient and generating corresponding microphone signals. A sound object identification module is configured for analyzing the microphone signals to identify one or more sound objects within the sound environment. A sound object selection module is configured for processing the microphone signals to generate a sound object signal for each of the one or more sound objects. A stimulation side selector module is configured for selecting on which side or sides of the bilateral cochlear implant arrangement to process each sound object signal. One or more sound processors processes the sound object signals to generate stimulation signals to the implanted electrode arrays on the selected side or sides.