基于脉搏波传导时间的无创连续血压测量研究

发布时间：2018-05-11 09:04

本文选题：血压连续测量 + 脉搏波传导时间　；参考：《东北大学》2012年硕士论文

【摘要】：血压是反映心血管功能的重要生理参数,是诊断疾病、观察治疗效果和进行预后判断的重要依据。相对比于传统的血压测量方法,无创连续血压测量方法可以摆脱气囊束缚,实现较长时间内血压情况的实时监测。无创连续血压测量方法有效地克服了传统血压测量方法的缺点,对于特殊情况下血压的监测以及临床监护具有一定的优势。在临床医学上,对危重病人和手术中的重症患者需要进行连续的血压监护,以便当病人出现意外时,医护人员能够及时采取救护措施。从以上分析看,连续血压测量在临床诊断中均具有重要意义。本文的研究基础为动脉血压和脉搏波传播的相关理论。通过改变体位得到多组变化血压值,建立了两种不同的利用脉搏波传导时间测量血压的数学模型,用袖带法测量的血压值与模型估算的血压值进行误差对比分析。本文主要完成了以下工作： (1)分析了数据采集过程中长时间采集和不同指端采集对脉搏波传导时间的影响。 (2)对采集到的信号进行预处理,将小波变换,数字滤波器,平均加权等去噪方法进行比较,确定去噪和去基线的最合适方法。 (3)将预处理后的ECG和脉搏波信号进行特征参数提取。用墨西哥小帽小波结合模极大值的方法确定ECG信号的R波,检测率达97%以上。使用差分阈值的方法确定脉搏波信号的波峰,检测率达98%以上。 (4)建立两种不同的利用脉搏波传导时间估算血压的模型。模型参数均为2个,模型1中的系数通过一次标定可得。模型2中的两个系数是通过多次测量的血压和脉搏波传导时间进行非线性回归得到。 (5)设计实验,分别标定两个模型,并且利用模型估算血压与袖带法得到的测量血压进行比较。结果两种模型血压估算精度均符合AAMI,模型1血压测量误差小于3.5±7.15mmHg,模型2血压测量误差小于2.58±6.64mmHg。
[Abstract]:Blood pressure is an important physiological parameter reflecting the cardiovascular function. It is an important basis for diagnosing the disease, observing the effect of the treatment and judging the prognosis. Relative to the traditional blood pressure measurement method, the non-invasive continuous blood pressure measurement method can get rid of the bondage of the air bag and realize the real time monitoring of the blood pressure in a long time. It has effectively overcome the shortcomings of the traditional blood pressure measurement methods, and has some advantages for the monitoring of blood pressure in special cases and clinical monitoring. In clinical medicine, continuous blood pressure monitoring is required for critically ill patients and severe patients in the operation. From the above analysis, continuous blood pressure measurement is of great importance in clinical diagnosis.
The basis of this study is the theory of arterial blood pressure and pulse wave propagation. By changing the position of the body to get a number of changes in blood pressure, two different mathematical models for measuring blood pressure using pulse wave conduction time are established. The error of the blood pressure measured by the cuff method is compared with the value of the model estimated by the model.
This article mainly completed the following work:
(1) the effects of long time acquisition and different fingertip collection on pulse wave propagation time were analyzed.
(2) to preprocess the collected signals, compare the denoising methods such as wavelet transform, digital filter and average weighting, and determine the most suitable method for denoising and removing the baseline.
(3) the characteristic parameters of the pre processed ECG and pulse wave signals are extracted. The R wave of the ECG signal is determined by the method of combining the modulus maximum of the Mexico small cap wavelet. The detection rate is over 97%. The peak of the pulse wave signal is determined by the method of differential threshold, and the detection rate is over 98%.
(4) to establish two different models of blood pressure estimation using pulse wave conduction time. The model parameters are all 2, and the coefficients in model 1 can be obtained by one time calibration. The two coefficients in model 2 are obtained by nonlinear regression through multiple measured blood pressure and Mai Bobo conduction time.
(5) the experiment was designed, two models were calibrated respectively, and the blood pressure was compared with the measured blood pressure obtained by the cuff method. Results the accuracy of the blood pressure estimation in the two models were all in accordance with AAMI, the model 1 blood pressure measurement error was less than 3.5 + 7.15mmHg, and the blood pressure measurement error of model 2 was less than 2.58 + 6.64mmHg..

【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R318.04

【参考文献】