Latency as a region contrast: Measuring ERP latency differences with Dynamic Time Warping

Area of Science:

• Neuroscience
• Signal Processing
• Biomedical Engineering

Background:

• Accurate measurement of onset latency is crucial for understanding neural signal processing.
• Existing methods for latency measurement can be sensitive to noise and parameter choices.
• A novel approach is needed to enhance the reliability and robustness of latency measurements.

Purpose of the Study:

• To evaluate a new dynamic time warping (DTW) algorithm for measuring onset latency contrasts.
• To compare the performance of DTW against traditional methods using computer simulations.
• To assess the sensitivity of DTW to noise, window size, and averaging techniques.

Main Methods:

• Computer simulations were used to compare DTW with existing latency measurement techniques.
• The study analyzed power and Type I error rates under various signal-to-noise ratios and window sizes (broad vs. narrow).
• Per-participant and group-level analyses were conducted using single-participant and jackknife average waveforms.

Main Results:

• The dynamic time warping (DTW) algorithm demonstrated superior performance compared to other evaluated methods.
• DTW exhibited reduced sensitivity to varying signal-to-noise ratios.
• The method proved less affected by the placement and width of the selected analysis window.

Conclusions:

• Dynamic time warping (DTW) offers a more robust and reliable method for measuring onset latency contrasts.
• This approach enhances the accuracy of latency measurements in the presence of noise and varying analytical parameters.
• DTW provides a valuable tool for neurophysiological research requiring precise signal timing analysis.