Digital Hearing Aids

ABOUT THE BOOK

The first book available on the subject, Digital Hearing Aids provides an overview of the signal-processing strategies implemented in modern digital hearing aids. Algorithms ranging from dynamic-range compression and directional microphones to sound classification and binaural noise suppression are clearly explained. The basic equations describing the signal-processing algorithms are presented along with full explanations for those less comfortable with the mathematics, and each processing strategy is accompanied by a summary of its effectiveness. The text is intended for a graduate audiology course in hearing aids and hearing-aid technology.

CONTENTS

Preface

1 .Hearing-Aid Technology
Types of Hearing Aids
Canal Aids
In-the-Ear Aids
Behind-the-Ear Aids
Open Fitting BTE Aids
Body-Worn Aids
From Analog to Digital
Moore’s Law
Digital Current Drain
Analog and Digital Power Consumption
Digital Circuit Components
Analog-to-Digital Converter
Digital Signal Processor
Memory
Clock
Batteries
Concluding Remarks
References

2. Signal Processing Basics
Signal and System Properties
Sequences
Linear Time Invariance
Convolution
Correlation
Transfer Functions
Feedback
Group Delay
Amplitude Modulation
Distortion
Discrete Fourier Transform
Filters and Filter Banks
Filter Definitions
Recursive Filters
Non-Recursive Filters
Frequency Resolution and Group Delay
Block Processing
Digital System Concerns
Integer Arithmetic
Quantization Noise
Aliasing Distortion
Temporal Aliasing
Algorithm Complexity
Algorithm Interaction
Concluding Remarks
References

3. The Electroacoustic System
Hearing Aid System
Head and Ear
Microphone and Receiver
Vent Acoustics
Acoustic Elements
Vent Size
Residual Ear-Canal Volume
Open BTE Fitting
Occlusion Effect
Concluding Remarks
References

4. Directional Microphones
Hearing-Aid Microphones
Microphone Construction
Directional Microphones and Spatial Signal Processing
Directional Response Patterns
Microphone Spacing and Time Delay
Microphone Mismatch
Frequency Response
Magnitude Frequency Response
Microphone Mismatch
Interaction with Vents
Microphone Noise
Microphones on the Head
Microphone Performance Indices
Front-to-Back Ratio
Directivity Index
Unidirectional Index The AI-DI
Rooms and Reverberation
Critical Distance
Early Reflections
Combining Noise and Reverberation
Benefit in the Real World
Concluding Remarks
References

5. Adaptive and Multi-Microphone Arrays
Two-Microphone Adaptive Array
Adaptive Delay
Adaptive Gain
Adaptive Filter
Delay-And-Sum Beamforming
Adaptive Arrays
Superdirective Arrays
Widely-Spaced Arrays
Array Benefits
Concluding Remarks
References

6. Wind Noise
Turbulence
Hearing-Aid Measurements
Air Flow and Turbulence
Sound Pressure
Signal Characteristics
Spectrum
Temporal Fluctuations
Correlation
Wind-Noise Reduction
Directional Microphones
Wind Screens
Spectrum Algorithms
Correlation Algorithms
Adaptive Arrays
Concluding Remarks
References

7. Feedback Cancellation
The Feedback System
System Equations
System Response
Gain-Reduction Solutions
Adaptive Feedback Cancellation
System Equations
LMS Adaptation
Constrained Adaptation
Adaptation Equations
Initialization
Simulation Results
Decorrelation Techniques
Interrupted Adaptation
Delay
Filtered-X Algorithm
Frequency-Domain Adaptation
Processing Limitations
Room Reflections
Measurement Procedure
Measurement Results
Maximum Stable Gain
Non-Linear Distortion
Concluding Remarks
References

8. Dynamic-Range Compression
Does Compression Help?
Algorithm Design Concerns
Frequency Resolution
Processing Delay
Single-Channel Compression
Input/Output Rules
Volume Control
Envelope Detection
Multi-Channel Compression
Temporal Response
Swept Frequency Response
Frequency-Domain Compression
Ideal FFT System
Practical FFT System
Side-Branch Structure
Frequency Warping
Digital Frequency Warping
Warped Compressor System
System Delay Comparison
Concluding Remarks
References

9. Single-Microphone Noise Suppression
Properties of Speech and Noise Signals
Low-Level Expansion
Envelope Valley Tracking
Bandwidth Reduction
Adaptive High-Pass Filter
Adaptive Low-Pass Filter
“Zeta Noise Blocker”
Envelope Modulation Filters
Concluding Remarks
References
Spectral Subtraction
Noise Estimation
Valley Detection
Minima Statistics
Histogram
Wiener Filter

10. Spectral Subtraction
Classical Approaches
General Equation
Nonlinear Expansion
Ephraim-Malah Algorithm
Auditory Masking
A Fundamental Compromise
Algorithm Effectiveness
Concluding Remarks
References

11. Spectral Contrast Enhancement
Auditory Filters in the Damaged Cochlea
Physiological Measurements
Tuning Curves
Neural Firing Patterns
Perceptual Measurements
Processing Strategies
Spectral Valley Suppression
Spectral Contrast Modification
Raise Spectrum to a Power
Spectral Filtering
Excess Upward Spread of Masking
F2/F1 Ratio
Processing Comparison
Combining Spectral Contrast Enhancement with Compression
Concluding Remarks
References

12. Sound Classification
The Rationale for Classification
Signal Features
Feature Selection
Mutual Information
Feature Selection Using Mutual Information
Classifier Algorithms
Training the Classifier
Types of Classifiers
Clustering
Linear Discriminant
Support Vector Machine
Neural Network
Hidden Markov Model
Bayes Classifier
Personalizing the Classifier
Classification Examples
Number of Features
Isolated Signals vs. Mixtures
Concluding Remarks
References
Binaural Signal Processing
The “Cocktail Party” Problem
Signal Transmission

13. Binaural Compression
Compression Using Binaural Loudness Summation
Auditory Efferents
Compression Using Auditory Efferents
Binaural Noise Suppression
Interaural Cross-Correlation
Binaural Wiener Filter
Binaural Signal Difference
Binaural Spectral Subtraction
Directional Cues
Localization Model
Dichotic Band Splitting
Concluding Remarks
References

Index

ABOUT THE AUTHOR

James M. Kates, E.E. holds the position of Research Fellow with hearing-aid manufacturer GN ReSound. He is also an Adjunct Professor in the Department of Speech Language and Hearing Sciences at University of Colorado in Boulder. Prior to moving to Colorado, he was Research Scientist in the laboratory of Harry Levitt at the City University of New York. He received the BSEE and MSEE degrees from the Massachusetts Institute of Technology in 1971, and the professional degree of Electrical Engineer from M.I.T. in 1972. He has developed many of the signal-processing algorithms implemented in the GN family of hearing aids, including the DFS feedback cancellation algorithm and the Warp-17 frequency-warped dynamic-range compression algorithm. He is a Fellow of the Acoustical Society of America and a Fellow of the Audio Engineering Society, and is the author or co-author of 47 technical papers and holds 17 US patents with seven patents pending. A widower, he resides in Niwot, Colorado.

AUDIENCE
Primary: Audiology