Neural Recording Circuit With Continuous Electrode Offset Cancellation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high-density neural implant systems face challenges in integrating large numbers of channels due to trade-offs between area, noise, power, and offset blocking capabilities, particularly with thermal and flicker noise, and existing offset cancellation methods are inefficient and prone to saturation.
Innovation Solution
A CMOS-based integrated circuit with a digitally assisted neural recording system that multiplexes multiple channels into a single amplifier, using a feedback loop and LMS adaptive filtering to simultaneously remove electrode offset voltages in the analog domain, employing a 2nd order ΔΣ modulator and digital signal processing to achieve continuous offset cancellation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple channels are integrated on a fully integrated implantable SoC, then the channel density increases, but the chip area increases and trades off with noise, power, and offset blocking capabilities
Solution Approach 1:
The patent merges multiple channel signals into a single shared amplifier through an analog multiplexer, allowing 16 channels to be recorded using one amplifier instead of 16 separate amplifiers. This significantly reduces the chip area required while maintaining high channel density, directly resolving the contradiction between channel quantity and chip area.
Solution Approach 2:
The patent segments the signal processing function into two parts: analog multiplexing for channel consolidation and digital demultiplexing for channel separation. This segmentation allows efficient resource sharing in the analog domain while maintaining full channel independence in the digital domain, enabling high channel density with reduced area.
2Area of stationary object
If a shared amplifier is used for multiple channels via TDMA, then the chip area is reduced, but the electrode offset voltage saturates the amplifier
Solution Approach 1:
The patent applies preliminary action by continuously estimating and canceling electrode offset voltages before they can saturate the amplifier. The offset cancellation operates proactively throughout the recording process, preventing saturation rather than correcting it afterward, ensuring reliable amplifier operation throughout the recording process.
Solution Approach 2:
The patent implements feedback by continuously monitoring the electrode offset voltage and using this information to adjust the cancellation signal in real-time. The estimated offset is fed back through a DAC to the amplifier input, creating a closed-loop system that actively prevents saturation and maintains reliable operation.
3Device complexity
If conventional offset cancellation using look-up table and periodic operations is used, then the implementation is simpler, but the offset cancellation is not continuous and is prone to saturation
Solution Approach 1:
The patent achieves continuous offset cancellation by continuously estimating the offset voltage and continuously applying the cancellation signal throughout the entire recording process. This eliminates the periodic interruptions of conventional methods, ensuring the amplifier operates reliably without saturation at any point during recording.
Solution Approach 2:
The system performs self-service by automatically estimating and canceling its own offset voltages without requiring external intervention or complex calibration procedures. The continuous adaptive offset cancellation operates autonomously, adjusting to changing offset conditions in real-time while maintaining simple implementation.
4Object-affected harmful factors
If large-area input differential pair amplifier is used to resolve flicker noise, then the flicker noise is reduced, but the chip area increases
Solution Approach 1:
The patent extracts the flicker noise problem from the amplifier design by using AC coupling to block DC offset and low-frequency flicker noise before they can affect the neural signal recording. This allows the use of a compact amplifier without requiring large-area input differential pairs, as the harmful low-frequency noise is removed through filtering rather than amplifier design.
Data Source
AI summary
A method for obtaining neural signals from a neural signal sensor includes extracting sensor offset from a neural input signal. Sensor offset is removed from N-channels that share a single amplifier. The neural input signal is acquired from the analog domain by neural electrodes with varying DC-offset and the method includes multiplexing the N-channels via a time divisional multiple access procedure into a single neural amplifier An integrated digitally assisted neural recording system includes an analog multiplexer structured to receive N-channels and to multiplex the N-Channels into a single neural amplifier. A feedback loop is configured to cancel electrode offset voltage in the digital domain by generating electrode offset voltage samples and adding delay to align each electrode offset voltage sample to be subtracted at the neural amplifier in the analog domain.