Summary:
UCLA researchers in the Department of Electrical and Computer Engineering have developed a novel hard-ware based system for real-time removal of stimulation artifacts in neural recordings, enabling greater scalability, flexibility, and efficiency in neuromodulation and neurotechnology applications.
Background:
Neuromodulation research depends on accurately measuring neural responses during electrical stimulation. High density electrode arrays enable these measurements but often capture strong stimulation artifacts that obscure real-time neural signals and hinder analysis. Existing methods to reduce these artifacts include temporal subtraction or spectral filtering implemented in recording amplifiers to extract underlying neural signals. Yet, both methods require dedicated hardware per channel, which limits scalability for larger electrode arrays. Additionally, in cases of randomization, biomimetic, or adaptive stimulation protocols these methods are unable to remove artifacts due to time-domain and frequency-domain characteristics. Software-based methods similarly introduce latency and suffer from limited dynamic range, resulting in signal loss. Thus, there is an unmet need for a scalable, real-time method to suppress stimulation artifacts and preserve high-fidelity neural recordings to advance developments in next-generation prosthetic devices and biomedical implants.
Innovation:
To address these limitations, Wentai Liu and his research team have developed a scalable, low-power multi-electrode neural acquisition system that removes stimulation artifacts directly at the amplifier stage, preventing saturation and signal loss. The proposed apparatus supports arbitrary stimulation protocols and offers high dynamic range, channel scalability, and energy efficiency. By performing artifact cancellation in real time through a dual-stage process, the system avoids common pitfalls such as quantization errors, digital processing delays, and the need for large memory buffers. Adaptive adjustments enable artifact suppression during and after stimulation, allowing simultaneous recording of immediate neural responses. This innovation has the potential to significantly advance neuromodulation research by enhancing the real-time signal-to-noise ratio, enabling access to neural signals that were previously hidden by stimulation artifacts. This improved data fidelity lays the groundwork for both therapeutic and technological breakthroughs. The system’s scalability and precision make it ideally suited for a range of applications, including neural prosthetics, brain-machine interfaces (BMI), neuromodulation therapies, and high-resolution neural mapping.
Potential Applications:
● Neuromodulation Therapies
● Brain-Machine Interfaces (BMI)
● Neuroprosthetics with sensory feedback
● High-resolution neuroscience research
● Intraoperative neural mapping and monitoring
Advantages:
● Artifact-free neural signals in real time
● Low memory and processing overhead
● Scalable to high-channel-count arrays
● Simultaneous stimulation and recording
● Compatible with arbitrary and adaptive stimulation protocols
● Enhanced signal-to-noise ratio
State of Development:
First successful demonstration of partial invention 04/12/23
Reference:
UCLA Case No. 2024-010
Lead Inventor:
Wentai Liu, Distinguished Professor, Bioengineering, Electrical and Computer Engineering