体感诱发电位术中脊髓功能监护异常逻辑判别准则的研究

Spinal surgery is an effective way of correcting deformities of the spine, but also entails the risk of damage to the nerveous system and complications caused by instrumentation compression to the spinal cord. The use of spinal cord monitoring in non-invasive, fast and effective way is necessary to prevent spinal cord nerve damage and avoid complications. In recent years, somatosensory evoked potential(SEP) is a widely used technique in spinal cord function monitoring and has become development direction of inraoperative spinal cord monitoring. Current SEP detection is limited to measure the amplitude and latency of time-domain waveform manually. However, SEP signals recorded in the operating theatre are usually accompanied by noise, which precludes the accuracy of detection by manual measurement, and lack of intelligent judgment auxiliary in monitoting also prevents the clinical application of intraoperative SEP monitoring. This project aims to employ time-frequency analysis and fuzzy neural network technology to research the methods of intraoperative SEP spinal cord function monitoring, extract new time-frequency parameters response to spinal cord injury more accurately,sensitively and stablly. To establish fuzzy logic decision combines with SEP time-frequency features, evoke potential impact factors and expert knowledge, achieve reliable and intelligent monitoring of inraoperative spinal cord function, solve the problem of poor reliability of spinal surgey and low degree of intilligence, provide a scientific basis for neurosurgery monitoring level and promote its clinical application.

脊柱外科手术是治疗各种脊柱脊髓病变的有效手段。但术中器械植入容易压迫脊髓、损伤神经系统和诱发严重并发症。为预防脊髓神经损害、避免并发症产生，十分必要进行无创、快速、有效的术中脊髓神经功能监护。近年来，体感诱发电位(SEP)是脊髓神经功能监护中最常用的电生理技术，成为术中脊髓监护技术的发展方向。但目前SEP检测仅限于人工测量时域波形幅值和潜伏期；由于术中存在强电磁干扰，人工测量难以保证检测精度，加之监护缺乏智能判别辅助，严重阻碍了术中SEP脊髓监护的临床应用。项目拟利用时频分析和模糊神经网络技术研究术中SEP脊髓功能监护方法，提取新的能更精确、灵敏、稳定反映脊髓损伤的SEP时频特征信息，创建融合SEP时频特征、诱发电位影响因素和专家知识的模糊逻辑决策器，实现术中脊髓功能的可靠智能监护，解决脊柱外科手术监护可靠性差、智能化程度低的问题，为提高神经外科手术监护水平、推动其临床应用提供科学依据。

体感诱发电位（SEP）是术中脊髓神经功能监护的重要手段，可以有效预防脊髓神经损害，避免发生并发症。目前，临床上SEP检测以静态基线为标准，需要人工测量SEP监护参数来判别脊髓功能状态，可靠性有待进一步提高。本项目创新性地提出了术中脊髓监护SEP异常动态预警判别模型，为术中脊髓监护提供了一种新方法。.在获得项目资助后的这三年里，我们完成了如下研究任务：开展了SEP时频分析, 统计分析了术中SEP监护的时-频域监护指标及监护标准，研究了脊髓急、慢性损伤前后SEP的主、子-时频成分变化，以及不同脊髓节段损伤后SEP时频成分的差异，揭示了SEP对脊髓损伤部位的诊断作用；通过对限制性二阶盲辨识算法的优化进行SEP的单次提取，分析了脊髓不同损伤阶段单次SEP在时域、时-频域的变异性，单次SEP时频特征能够很好地反映脊髓损伤变化过程中的脊髓功能变化状态，不仅体现了术中SEP的动态变异信息，而且其时频特征参数具有更好的稳定性。在脊髓损伤过程中单次SEP在时域、时-频域的变异性特征具有一致性；深入研究了脊髓损伤前后单次SEP时频成分变化，及不同脊髓节段损伤后单次SEP时频成分的差异，进一步揭示了单次SEP的时频信息能更好地确定脊髓损伤部位；重点研究了麻醉对术中SEP的影响，分析了麻醉剂对术中诱发电位的影响规律，并确定了影响术中SEP变异性的主要因素，将有助于正确解释术中诱发电位监测的结果，提供可靠的术中神经功能监护；建立了综合多因素的术中脊髓监护动态SEP异常预警判别模型，对术中SEP异常进行逻辑判别，能够更客观的对脊髓功能状态进行监护。.本项目提出新型的综合参数异常预警判别模型，充分体现了脊髓神经功能的动态变化，与传统潜伏期延长10%或者幅值减低50%的SEP异常预警标准相比，降低了外科手术监护过程中对监护人员的专业经验要求，避免了监护专家测量SEP主观性对监护结果的影响，提高了术中监护结果的可靠性。