Asynchronous Event-driven Encoder With Simultaneous Temporal Envelope and Phase Extraction for Cochlear Implants.

Area of Science:

• Biomedical Engineering
• Electrical Engineering
• Signal Processing

Background:

• Conventional cochlear implants (CIs) suffer from high power consumption and limitations in capturing acoustic signal amplitude and phase.
• Periodic sampling methods in CIs are inefficient for extracting the temporal fine structure of sound.

Purpose of the Study:

• To present an asynchronous event-driven encoder chip for cochlear implants.
• To enable simultaneous extraction of amplitude and phase information from acoustic signals.
• To reduce power consumption and circuit area in cochlear implant technology.

Main Methods:

• Developed an asynchronous delta modulation (ADM) based chip architecture.
• Captured and digitized signal peak/trough crossing events intrinsically.
• Fabricated an 8-channel prototype chip in a 0.18 μm 1P6M CMOS process.
• Integrated two chips to create a 16-channel stimulation encoding system.

Main Results:

• The 8-channel prototype chip occupies a small area (0.125 × 1.7 mm²) and consumes low power (36.2 μW at 0.6V).
• The 16-channel system effectively processes the full audible frequency range (100 Hz to 10 kHz).
• ADM architecture intrinsically reduces power consumption, circuit area, and eliminates the need for data compression.

Conclusions:

• The asynchronous event-driven encoder chip offers a power-efficient solution for cochlear implants.
• This technology enables superior extraction of temporal fine structure, crucial for speech perception.
• Experimental characterization confirms the chip's functionality for real-world auditory applications.