非接触式颅内疾病动态变化特征提取方法研究

Along with people life level of exaltation，the incidence of cerebrovascular disease increased year by year，it has gradually become the most common Encephalopathy. If there is no prompt treatment, the death rate and disability rate are very high undoubtedly. Early, timely and accurate evaluation of, the location, nature and extent or severity, as well as the development , this is the key to the neurological emergency rescue. The traditional detection technology, e.g., head CT, Magnetic Resonance Imaging (MRI), can accurately determine the critical nature, range and degree of the disease, but these aren’t still available to continuous imaging for monitoring. However in some critical cases, patients should not be moved repeatedly，and they cannot be monitoring dynamic changes of the lesions using these devices. Hence, these existing devices are limited to the timely judging disease and adjusting treatment. The project regards intracranial disease of hematoma and intracranial pressure in the process of the lesions as the research object, see magnetic induction imaging and ultrasonic testing technology research as the basis, and study the simulation model of the dynamic variation of the hematoma and the process of intracranial pressure change. At the same time ,optimize extracting dynamic feature algorithm of magnetic induction tomography and ultrasonic testing technology. Combining the above two techniques, developed a non-contact, small, cheap real-time continuous monitoring of brain hematoma and dynamic changes of intracranial pressure equipment.

随着物质生活水平的改善，脑血管病的发病率逐年上升，脑血管病逐渐成为神经内科最常见疾病，若不能得到及时的救治，其致死率及致残率均很高。如果临床上能够早期、及时准确评价各种病因所致脑组织病变的部位、性质及范围或严重程度，以及其发展，这是神经科急危重症患者抢救成败的关键。虽然目前的检测技术如CT、核磁共振（MRI）等，可以准确判定危重疾病的性质、范围和程度,但是无法进行床旁连续的成像，对于病情危重、变化较快，但又不宜反复搬动的患者就不能监测其病灶的动态变化，因此对病情演变的及时判断及调整治疗方案受到限制。本项目以颅内疾病过程中血肿和颅内压为研究对象，以磁感应成像技术和超声波检测技术研究为基础，系统研究了模拟血肿动态变化的模型和模拟颅内压变化过程的模型，优化磁感应成像技术和超声波检测技术的动态特征提取算法，结合以上两种技术，研制出一种非接触、小型、廉价的实时连续监测脑血肿和颅内压的动态变化的设备。

本项目以颅内疾病过程中血肿和颅内压为研究对象，以磁感应成像技术和超声波检测技术研究为基础，系统研究了模拟血肿动态变化的模型和模拟颅内压变化过程的模型，优化磁感应成像技术和超声波检测技术的动态特征提取算法，结合以上两种技术。在前期主要研究人脑模型结构设计，在此基础上，每次按5mL的量注入模拟的血液，探索血肿体积大小的变化，引起磁感应信号和超声信号的变化情况。结果表明，血肿变化到50-60mL，磁感应相位变化量有200mDeg变化量，超声检测信号相位最大的变化差值有10Deg，通过相位能够完全区分出5mL出血量。针对测量系统随机漂移的问题，采用相位漂移动态抑制电路，同时改进锁相检测电路，保证输出相位信号的稳定性。针对成像逆问题的分辨率差，在重构算法上，提出导数法峰值锐化处理测线圈的测量数据，提高多目标分辨率。将双目标成像结果的相关系数增加了15%，归一化均方距离（Normalization Mean Square Distance criterion, NMSD）和归一化平均绝对距离（Normalization Mean Absolute Distance criterion, NMAD）误差减小了约0.5。将原子磁力计引入磁感应成像系统，研究系统锁频技术，提高系统稳定性。研制出一种非接触、小型、廉价的实时连续监测脑血肿和颅内压的动态变化信息的样机，为项目走向临床打下基础。