Setting adaptive spike detection threshold for smoothed TEO based on robust statistics theory.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Extracellular neural recordings are crucial for understanding brain activity.
• Accurate spike detection in neural signals is essential for diagnostics.
• Existing methods for adaptive thresholding in spike detection have limitations.

Purpose of the Study:

• To propose a novel adaptive thresholding method for the smoothed Teager energy operator (STEO) detector.
• To enhance the accuracy and robustness of spike detection in neural and other biomedical signals.
• To provide a method that does not require prior knowledge of neural spike waveform shapes.

Main Methods:

• Derived the relationship between input and output signal statistics for the STEO detector.
• Utilized robust statistics theory for unbiased estimation of background noise statistics.
• Validated the method using synthetic and real extracellular neural recordings from various sources (monkey, rat, human).

Main Results:

• The proposed adaptive thresholding method effectively detects neural spikes.
• The approach demonstrated robustness across diverse neural recording datasets.
• Simulation results indicate superior performance compared to existing state-of-the-art adaptive detection methods.

Conclusions:

• The novel adaptive thresholding approach for STEO detectors is effective and robust for neural signal analysis.
• This method offers a significant improvement over current adaptive detection techniques.
• The approach is broadly applicable to biomedical signals where spike detection is critical.