Automated detection of spikes in EEG is a current research focus and is important for epilepsy diagnosis. There are two main types of existing detection methods: signal analysis and machine learning. The former is sensitive to outliers, and the robustness of the latter algorithms to different data has not been fully validated. In addition, most of the current studies are based on single-channel EEG, which is usually susceptible to artefactual interference. Aiming at the problems of the existing algorithms and combining the characteristics of spikes, this paper proposes a spike detection algorithm based on multi-channel data weight fusion and wavelet decomposition, which adopts feature fusion based on the weights of multi-channel data to achieve data reinforcement of spikes, and finally performs wavelet decomposition of the data and detects the spikes using the mode maxima. Experimentally verified, the algorithm achieves the precise detection of interictal spike wave, and the diagnostic accuracy can reach more than 92.3%, which provides a informative method for the detection of epileptic spike wave.