应用连续小波变换提取脊柱手术中脊髓功能特征参数的研究

Application of CWT to Extract Characteristic Monitoring Parameters During Spine Surgery

在脊柱外科手术中,为防止脊髓神经功能损伤进行术中脊髓功能监护十分必要。本研究以特发性脊柱侧弯矫形手术中脊髓功能的电生理监护特征参数提取技术为目标,试图解决监护参数在时域参数变异性较大、容易受干扰的问题。运用连续小波变换方法对手术中采集的体感诱发电位分析处理,提取信号在时频域的三个新的特征参数。通过本研究表明,在脊柱手术监护中,运用连续小波变换获得的时频域特征参数比单纯在时域分析参数具有更小的变异系数。因此时频域特征分析参数比单纯时域的幅值和潜伏期分析更稳定和可靠。在脊柱手术中如果应用时频监护参数可能会更有效地避免脊髓损伤。

It is necessary to monitor intraoperative spinal function in order to prevent spinal neurological deficit during spine surgery. This study aims to extract characteristic electrophysiological monitoring parameters during surgical treatment of scoliosis. The problem, “the monitoring parameters in time domain are of great variability and are sensitive to noise”, may also be solved in this study. By use of continuous wavelet transform to analyze the intraoperative cortical somatosensory evoked potential （CSEP）, three new characteristic monitoring parameters in time-frequency domain （TFD） are extracted. The results indicate that the variability of CSEP characteristic parameters in TFD is lower than the variability of those in time domain. Therefore, the TFD characteristic monitoring parameters are more stable and reliable parameters of latency and amplitude in time domain. The application of TFD monitoring parameters during spine surgery may avoid spinal injury effectively.