磁感应相位移谱用于脑出血时颅内主要容积物变化监测的研究

Intracerebral hemorrhage (ICH) is the second major types of stroke and the monitoring of ICH is a key step in the treatment for patients. The magnetic induction brain monitoring technology is based on the electrical properties of brain tissues. By magnetic induction technology, ICH can be noninvasively and continuously monitored beside the bed with non-contact. But the technology treats the brain as a whole during monitoring currently. Its monitoring signal is absence of a physiological explanation combined with development of ICH and it has not been developed to monitor the intracranial physiological changes along with ICH development. Previously we have verified that the ICH magnetic induction monitoring signal was caused by changes of main tissues in brain and the signal contained information about changes of the tissues. Based on our previous research results, this study proposes that utilizes the magnetic induction phase shift spectroscopy (MIPSS) to monitor changes of main brain tissues in ICH, and advances the magnetic induction technology from whole brain monitoring to the main brain tissues monitoring. It uses the bama miniature pig as the research object. The study utilizes medical imaging and developed home-made image processing algorithms to quantify the volume of cerebrospinal fluid (CSF) changes and cerebral blood flow (CBF) changes. And it illustrates the influence process of the main brain tissues on the MIPSS signal in ICH monitoring by statistical method of binary logistic regression and correlation analysis, which is based on data of the quantified volume of CSF and CBF changes and the hematoma injected. The magnetic induction monitoring parameters are established corresponding to hematoma, CSF and CBF by MIPSS signal processing, feature extraction methods and least squares support vector regression method. This study will contribute to improve the practical value of magnetic induction technology on the ICH monitoring.

脑出血是脑卒中的第二大类型，其连续实时监测是患者预后和治疗成败的关键。磁感应技术基于组织电特性实现脑出血监测，可在床旁非接触连续监测。但当前该技术将大脑作为整体监测，结果缺乏结合病情发展的医学生理解释，难以监测出血进程中颅内病生理变化。课题组前期研究证实，磁感应监测信号是由颅内主要容积物变化共同引起，信号包含它们的变化信息。本课题提出利用磁感应相位移谱监测脑出血时颅内主要容积物变化，将磁感应脑监测由大脑整体推进到组织层面监测。课题以前期研究成果为基础，以巴马香猪脑出血模型为研究对象，利用医学成像和自备的图像处理方法，通过二元logistic回归和相关偏相关分析等手段，阐明脑出血时颅内主要容积物对磁感应相位移谱信号的影响过程，结合磁感应信号的处理、特征提取及非线性回归建模等方法，建立反应血肿、脑脊液和脑血流变化的磁感应监测参数。课题的开展将为提高磁感应技术在脑出血监护中的应用价值建立基础。

脑出血病死率高，需要快速检测并及时采取医护措施，在脑出血病例中出血性脑卒中最为常见，此外颅脑外伤也是脑出血的常见致病原因。脑出血治疗前须经检测与评估，而当前临床上脑出血检测主要通过CT、MRI等影像学手段，但难以在床旁或院外开展，而基于生物电磁探测原理的便携快速脑检测技术是一种非常有前景的手段。.本课题，①开展了闭合性颅脑出血电磁检测研究。通过在特征频带上基于磁感应相位移方法建立了一种早期、非接触、无创的闭合性颅脑出血（Closed cerebral hemorrhage，CCH）检测系统，对20只CCH家兔进行实验研究。通过撞击仪对2组家兔建立不同程度的CCH模型，采用CT影像与病理学手段研究家兔在致伤后CCH的情况。结果表明轻型损伤组和重型损伤组反射参数S11、传输参数S21致伤前后平均相位移具有显著性差异。②提出一种基于宽带天线技术的脑出血检测方法，设计了一侧边接地的微带天线，对13只家兔进行了脑出血实验研究，并对家兔脑出血严重程度进行分类与预测。采用留一法进行交叉验证，显示这种方法能有效判别脑出血的严重程度（准确率77%）。③开展基于生物电磁检测原理的脑卒中类型检测鉴别研究，首先以不同浓度的生理盐水来分别模拟具有不同电导率的血液、脑脊液、白质和灰质，开展了物理模拟实验。再分别就单通道检测、多通道检测进行了动物实验和数据分析，特别基于多通道检测方法，对于脑出血1ml的鉴别准确率达到了80%，而对于脑缺血12min的鉴别准确率达到了85%，对脑出血和脑缺血进行鉴别，判定准确率达90%。.通过本课题研究，表明在闭合性脑出血检测中，电磁检测实验表明其可无创非接触判断颅内损伤程度，且传输参数要优于反射参数，为下一步的研究与应用建立了实验依据；基于宽带天线技术的脑出血检测系统具有对脑出血进行早期诊断的潜力，并有望对出血的严重程度进行区分；更表明了生物电磁脑检测技术在脑卒中类型的院前检测鉴别上具有极大应用价值，同时为生物电磁脑检测技术的进一步发展建立了基础。