A Novel Extracellular Spike Detection Algorithm Based on Improved Complete Ensemble Empirical Mode Decomposition with Adaptive Noise

Precisely discerning spike signals amidst noise presents a significant and intricate challenge. To address this issue, we propose a new detection algorithm for extracellular spikes based on improved complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN), aiming to achieve high accuracy and robust spike detection. Firstly, the extracellular recording signal is decomposed using CEEMDAN to obtain multiple intrinsic mode functions (IMFs). Subsequently, by analyzing the sample entropy of these IMFs, the important information hidden in the recorded signal can be prevented. Finally, the high-entropy noise components are discarded using wavelet soft thresholding. Experimental results from both simulated and real data demonstrate that the proposed method significantly outperforms existing methods by a large margin.