提出了一种基于小波变换进行心脏室颤波特征点识别的新方法。该方法首先对信号在特定不同尺度上的小波变换结果进行“多尺度交叉搜索” ,然后再在此基础上进行自适应判决检测。实践证明该方法的识别率达到 94 5% ,通过与常规检测方法进行比较发现它具有明显的优越性 A new method based on wavelet transform to identify feature points of ventricular fibrillation was proposed. This method first performs a “multi-scale cross-search” on the results of wavelet transforms at specific different scales, and then performs adaptive decision detection on this basis. Practice has proved that this method has a recognition rate of 945%, and it has obvious advantages through comparison with conventional detection methods.