連續血壓測量之訊號分析與實現

Signal Analysis and Implementation of Cuffless Continuous Blood Pressure Measurement

韓易(Yi Han)

指導教授 ： 林康平

中原大學/電機資訊學院/電機工程研究所/碩士(2016年)

https://doi.org/10.6840/cycu201600675

摘要

連續性無袖式血壓量測的在近年越來越受到重視，其量測方法主要是藉由血壓對於脈波傳導速度(PWV)關係的數學模型，而脈波傳導速度是由脈波傳輸時間(PTT)與血液流過的路徑所計算求得，所以最後只需考慮PTT對於血壓的關係，近年來有許多與血壓之變化具有相關性的參數不斷被發現，進而推導出許多新的基於PTT之血壓數學模型，目的是為了使估測的血壓值更準確，以及對於實際血壓值具有高度的相關性，除了改變血壓的模型外，有些方法會以脈波到達時間(PAT)取代PTT，PAT與PTT參數主要的差異在於，PTT有消除預射血前期(PEP)的時間，能夠更精確的計算脈波傳導速度(PWV)，但是基於PTT或基於PAT的估測準確性能是有爭議的。
本論文的實驗使用了6位受試者，僅考慮在休息的情況下所量測的心電圖與光體積變化描記圖，這兩種訊號的取樣頻率皆為2000Hz，其目的是為了減少估測系統所造成的計算誤差，並且由這兩種訊號的關係計算PAT與PTT。本論文使用了12種不同的血壓估測模型，這些模型分別為近年來所提出的6種模型，以及本論文修正後的6種新模型，這6種新模型主要是考慮兩兩脈波之間的時間間隔(PPI)能夠反映血壓的變化，為了探討這些模型在估測上的準確性，並比較了各模型的估測結果，此外也討論了基於PTT或基於PAT的估測結果，這些估測結果即為估測值與實際值的相關係數與誤差大小。
最後本論文的結論可分為四大部份，第一部份為使用越多參數的估測結果會呈現較高的相關性與較低的誤差，在估測的準確性上有較佳的結果，並且發現模型編號8具有最佳的估測準確性，其相關係數最高可達0.92，第二部份為使用PAT或PTT作為估測參數的比較結果，發現使用PAT的估測結果之相關係數最高，所以使用PAT作為估測參數會有最佳的估測準確性，第三部份為使用PPI修正後的模型會有較佳的估測準確性，其原因在於PPI較能反映出呼吸所造成的血壓變動的情形，第四部份討論了取樣頻率分別為2000Hz、500Hz與250Hz所造成的血壓估測系統上的誤差，發現取樣頻率越低造成的誤差會越大，所以本論文使用2000Hz的取樣頻率能夠減少血壓估測系統上的誤差。

關鍵字

脈波到達時間 ； 脈波傳輸時間 ； 脈波傳導速度 ； 連續性無袖式血壓量測

並列摘要

The continuous cuffless blood pressure measurement has been valued in recent years, and the measurement is based on a mathematical model of blood pressure (BP) to pulse wave velocity (PWV). PWV is calculated from pulse transit time (PTT) and the path of blood flow. A lot of parameters related to BP variation have been obtained, and different BP mathematical models based on PTT have been developed, by which the estimated BP could be more accurate. In the models, signals from one lead ECG and single potoplethysmography (PPG) were concerned to estimate BP so parameters from PTT might be replaced by Pulse Arrival Time (PAT). However, no matter the primary parameters from PAT or PTT, the accuracy was still required to be concerned.
6 subjects were involved with the study. Signals from ECG and PPG were obtained during resting, which sampling rates were both 2000Hz with the purpose to reduce calculation errors and to calculate PAT and PTT based on the correlation of the two signals. 12 BP estimating models were applied in the study; including 6 models having been developed during the recent years, and 6 new modified models, which parameters of PW were replaced by which of PPI because PW was inevitably affected by PPG on account of pulse to pulse interval (PPI) reflecting immediate BP variation. In order to discuss the effects of the 6 original models and the 6 new modified ones, when BP accuracy was estimated, in addition to compare the estimated results, results from PTT or PAT were discussed, and the related parameters and errors of the estimated value and the actual ones were evaluation index for the results.
There were four conclusions in this study. First, the estimated results with more parameters showed higher correlation and lower errors with better results in accuracy; and the 8th model had the optimal result in accuracy and its correlation coefficient reached 0.92. Second, PAT or PTT was applied as the comparative result of the estimated parameters, which showed that the application of PAT revealed the highest correlation coefficient, and the application of PAT brought the optimal estimated accuracy. Third, the model modified by PPI presented better estimated accuracy because PPI could reflect the BP variation from breathing. Forth, the sampling rate were 2000Hz、500Hz and 250Hz, and it was found that the lower sampling rate was, the more errors occurred. Therefore, sampling rate of 2000Hz was applied in this study, which could reduce errors of BP estimated system.

並列關鍵字

Pulse arrival time ； Pulse transit time ； Pulse wave velocity ； Continuous cuffless blood pressure measurement

參考文獻

[3] Saugel, B., Dueck, R., Wagner, J. Y. "Measurement of blood pressure. " Best Practice & Research Clinical Anaesthesiology 28.4 (2014): 309-322.

[4] Mozaffarian, Dariush, et al. "Executive Summary: Heart Disease and Stroke Statistics-2016 Update: A Report From the American Heart" Asso ciation.Circulation, 2016, 133.4: 447.

[5] World Health Organization, http://www.who.int/en/

[6] Welch Allyn "Blood Pressure Training"

[9] White, W. B., Berson, A. S., Robbins, C., Jamieson, M. J., Prisant, L. M., Roccella, E., & Sheps, S. G. "National standard for measurement of resting and ambulatory blood pressures with automated sphygmomanometers."Hypertension 21.4 (1993): 504-509.