Towards a Unified Framework for De-noising Neural Signals

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Neural signals are crucial for decision-making across various fields.
• Low signal-to-noise ratio (SNR) and artifacts, like motion artifacts in electroencephalography (EEG), complicate signal interpretation.
• Existing denoising tools are often limited to offline processing, lacking real-time capabilities.

Purpose of the Study:

• To develop high-performance, reliable, and real-time methods for neural signal artifact handling.
• To address the challenge of denoising complex artifacts, including EEG motion artifacts.
• To establish a unified framework for neural data artifact denoising compatible with real-time applications.

Main Methods:

• Development of novel sample-adaptive processing techniques.
• Application of a core processing tool to diverse artifact types.
• Focus on real-time processing capabilities for neural data.

Main Results:

• Proposed methods effectively handle challenging neural signal artifacts, including motion artifacts.
• Demonstrated the potential for a unified artifact denoising framework.
• Achieved real-time processing compatibility for neural data artifact removal.

Conclusions:

• Novel methods offer a promising solution for real-time neural signal artifact denoising.
• The unified framework approach simplifies and enhances artifact management in neural data.
• This work addresses a critical need for effective, real-time artifact removal in neuroscience and medicine.