Audio Signal Processing for Hearing Aid Speech Intelligibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern hearing aids face issues such as high cost, potential for additional hearing loss due to over-amplification, short battery life, feedback, aesthetics concerns, earwax accumulation, skin irritation, and ineffectiveness in noisy environments, particularly in improving speech intelligibility.
Innovation Solution
An audio system comprising a filtered volume determiner, fixed volume adder, filtered noise generator, signal modulator, and mixer that filters and amplifies specific frequency ranges to enhance speech intelligibility by converting unvoiced phone sounds into amplitude modulated noise, which is then mixed with the original signal to improve audibility and reduce noise interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional hearing aids amplify sound to improve speech intelligibility, then speech understanding improves, but additional hearing loss occurs due to over-amplification
Solution Approach 1:
The patent divides the audio signal into multiple frequency channels using filter banks, allowing independent processing of different frequency ranges. This segmentation enables precise control of amplification in each channel, preventing over-amplification in any single frequency band while maintaining overall speech intelligibility.
Solution Approach 2:
The patent applies different processing strategies to different frequency channels based on their specific characteristics and the user's hearing thresholds. Each channel can have customized gain, compression, and noise reduction parameters, providing locally optimized quality rather than uniform processing across all frequencies.
2Reliability
If hearing aids provide sufficient amplification for speech understanding, then speech intelligibility improves, but feedback and squealing noises occur
Solution Approach 1:
The patent extracts and removes feedback signals from the audio processing chain using adaptive algorithms. By identifying and subtracting the feedback component from the amplified signal, the system maintains sufficient gain for speech understanding while eliminating the harmful feedback and squealing noises.
Solution Approach 2:
The patent employs feedback control mechanisms where the output signal is continuously monitored and used to adjust the amplification in real-time. This closed-loop approach detects and compensates for feedback conditions, allowing the system to maintain optimal gain levels without producing squealing noises.
3Reliability
If hearing aids use recursive pickup and amplification to enhance speech, then speech intelligibility improves, but disruptive squealing noises are generated
Solution Approach 1:
The patent applies preliminary processing steps including noise reduction and signal enhancement before the recursive amplification stage. By pre-processing the signal to remove background noise and enhance speech components, the system reduces the likelihood of feedback and squealing during subsequent amplification cycles while maintaining speech intelligibility.
4Reliability
If hearing aids amplify all frequency ranges to improve speech understanding, then speech intelligibility improves, but noise interference increases
Solution Approach 1:
The patent applies noise reduction and signal enhancement techniques selectively to specific frequency channels where speech components are present, rather than uniformly processing all frequencies. This localized approach improves speech intelligibility in speech-containing channels while leaving noise-dominated channels unaffected, thereby reducing overall noise interference.
Solution Approach 2:
The patent dynamically adjusts processing parameters such as gain, compression ratio, and noise reduction thresholds based on the characteristics of each frequency channel and the detected signal type. By changing parameters adaptively, the system enhances speech components while suppressing noise, improving speech intelligibility without proportionally increasing noise interference.
Data Source
AI summary
In one embodiment, an audio system can replace a portion of an audio signal within a first range of frequencies, with an amplitude modulated noise signal comprising frequencies within the first range of frequencies and having a volume envelope corresponding to a volume envelope of the audio signal within a second range of frequencies.