Wavelet methods for spike detection in mouse renal sympathetic nerve activity.

Area of Science:

• Neuroscience
• Cardiovascular Research
• Biomedical Engineering

Background:

• Abnormal autonomic nerve activity is linked to peripheral neuropathies and cardiovascular diseases.
• Genetically altered mice are used to study these conditions, with sympathetic nerve activity recording as a key assessment method.
• Current methods for detecting murine sympathetic spikes rely on manual voltage thresholds, lacking unsupervised approaches.

Purpose of the Study:

• To develop and evaluate unsupervised spike detection algorithms for murine sympathetic nerve recordings.
• To compare the performance of wavelet-based methods against traditional amplitude discriminators.
• To identify the most robust algorithm for accurate spike detection in research settings.

Main Methods:

• Simulated murine renal sympathetic nerve recordings were used to test algorithms.
• Algorithms included automated amplitude discriminators and wavelet-based methods (Discrete Wavelet Transform - DWT, Stationary Wavelet Transform - SWT).
• Wavelet method parameters were optimized using basal, postmortem, and pharmacologically altered sympathetic activity recordings.

Main Results:

• Stationary Wavelet Transform (SWT) methods generally outperformed amplitude discriminators and Discrete Wavelet Transform (DWT) methods.
• A specific SWT method estimating noise levels and thresholding relevant signal scales demonstrated the most robust spike detection.
• The validated noise-level estimation method proved effective during pharmacological interventions.

Conclusions:

• Unsupervised wavelet-based methods, particularly SWT, offer a significant advancement over manual thresholding for murine sympathetic spike detection.
• The proposed noise-level estimation technique provides a reliable and validated approach for analyzing autonomic nerve activity in mice.
• This improved detection method will facilitate genetic and molecular studies of autonomic disorders.