节点文献

生命体征信号的穿戴式采集传输与智能处理

The Wearble Collection,transmission,and Intellegent Processing of Vital Signs

【作者】 林绍杰

【导师】 吴效明;

【作者基本信息】 华南理工大学 , 生物医学工程, 2010, 硕士

【摘要】 穿戴式医学检测技术是远程医疗及面向社区家庭的医疗服务系统中的重要基础,涉及生命体征信息获取、无线数据传输、医学信号处理、微机电系统MEMS及服务软件设计等多项关键技术。本文在深入分析无约束医疗监护的特点和技术要求的基础上,重点对生命体征信号检测模块的微型化、低功耗设计,基于ZigBee的无线信号传输网构建以及消除信号运动伪迹干扰的智能信号处理方法等关键技术进行了研究。基于硬件软件化的设计思想,将传统电路的高阶滤波、脉冲激励等部分由单片机甚至上位机来完成,体积缩小到传统模块的十分之一,解决了生命体征采集模块的微型化问题;分离检测模块与无线模块,实现了无线模块的互换性和通用性;设计电源管理电路解决无线传输与检测模块中大能耗电路的竞争问题,同时减少数字部分对模拟部分电路的干扰,提高信号质量。采用ZigBee技术构建躯域传感网,实现生理信号高可靠、低功耗的无线传输;设计缓冲技术以及数据请求/应答的数据传输模式,解决低速率下的高吞吐量数据传输的实时性和时间同步;测试网络在蓝牙环境及GSM环境下的抗干扰性,提出在医院病区病房中安全使用躯域传感网群体的措施。以统一过程的迭代开发理论为指导,以设计原则和设计模式为基础,用统一建模语言描述了整个无线监护基站软件的分析设计过程。包括对无线监护业务的领域建模,对各种用例的需求分析,进而设计出具有可靠性、可维护性、可复用性的软件系统。实现了信息管理、数据采集、信号处理、异常报警、数据管理、信号回放、网络通信等功能。最后,针对移动环境下生理信号的运动伪迹问题,提出同步采集心电和电极皮肤接触阻抗信号作自适应滤波消除运动伪迹噪声的方法。实验效果显示,采用120Hz的25μA激励电流,经过三阶RLS算法对心电信号进行实时修正,能够有效消除90%的基线漂移。另外,在分析各种临床心电波形的基础上,提出了严重心率异常报警算法,对MIT/BIH数据库中严重异常心电片断的检出率达到99.991%,能够有效预警临床各类短暂性且具有生命危险的心律失常。研究成果表明,穿戴式生命体征无线监护技术在实验环境下切实可行,有效提高了诊疗水平和病人的生活质量,将具有广泛的应用前景。

【Abstract】 Wearable vital signs monitoring is a key technology which can improve the efficiency of diagnosis and treatment for the convalescent and solve the headache of expensive medicine. Its core technology relates to the collection of vital signs, the wireless transmission of signals, the design of system software, and the intelligent processing of bio-signals. The aim of our project is to provide the patients with unconstrained and comfortable monitoring services and to provide the medical staffs with the efficient clinical information of patients at all times and places. The main content of this paper is to develop a wearable vital signs monitoring system and to study the key posers and their solutions in the development process. Analyzed on the principal of wearable vital signs monitoring technology and the feasibility of system design scheme, the paper presents the targets, the solutions and architecture of the system.Based on the Soft Hardware, many functions in traditional circuit, such as high order filter and pulse driver, are carried out by the Micro Controller Unit, or even the computer. Thus, the size of miniature module is reduced to one tenth of the traditional module. The separation between detection module and wireless transmission module improves the interchangeability and interoperability of the wireless module. Furthermore, the power management circuit copes with the power competition between the wireless module and the large energy consumption circuit in detection module.The body sensor networks, constructed via ZigBee, are responsible for the high reliability and low power combustion wireless transmission of physiological signals. Buffer technology and Data Request / Acknowledge mechanism are designed to realize the real-time high throughput data transmission and the time synchronization in low rate. The paper also shows the result of immunity of networks coexisting with Bluetooth or GSM and guides the safety usage of body sensor networks.Guided by the Iterative Development Theory in Unified Process and the design principles and design patterns, the paper describes the software design process of the entire wireless monitoring base station using the unified modeling language. Based on the business modeling of wireless monitoring and the requirement analysis of use case, a reliabe, maintainable, reusable software system is designed. At last, the paper focuses on the intelligent processing of bio-signals in ambulatory environment. On the one hand, the motion artifact of vital signs is cancelled by the methods for removing it from ECG signals via synchronously acquiring the elector/skin impedance as the reference signal of adaptive filter. The results demonstrated that 3rd order RLS adaptive filter can remove 90% of the baseline shift effectively with an exciting current of 120Hz and 25μA. On the other hand, to early warn the temporary life-threatening arrhythmia, an algorithm for alarming the seriously abnormal heart rate is presented. The detection rate of severely abnormal ECG segments is 99.991%.Above of all, the implementation of the system indicates that it is practicable and improves the diagnosis and treatment and patient quality of life with lower cost, which will have wide application.

节点文献中: