direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Inhalt des Dokuments

Advanced Continuous Blood Pressure Trend Estimation

What is the goal?

rBSN Biosignal Logo

The primary objective of this research initiative is to develop new methods of acquiring a continuous, unobstructive and accurate estimate of the blood pressure trend in humans. This entails the conceptualization, realization and validation of the hard- and software, as well as the signal processing required therefor. [1]

Why is it important?

ABP measurements on the arm
ABP measurements on the arm

Persistent and pathological impairment of blood pressure regulation leads to installation of hypo- or hypertension, strongly associated with cardiovascular risk. Vascular age is a fairly new concept, aimed at assessing the current cardiovascular health of an individual. An increased vascular age greatly heightens the risk of vascular disease and subsequent complications [2].

Recent studies reveal that more than 20% of Americans are affected by hypertension. More worryingly, up to 25% thereof are unaware of their condition [3]. Hypertension is characterized by increased systolic and diastolic blood pressure (>140 / 90 mmHg), and leads to heart and vascular hypertrophy, which can cause heart failure and vascular damage of organs.

Hypotension is mostly asymptomatic and has less severe consequences than hypertension. It is described by low systolic blood pressure (<100 mmHg) and therefore mild hypoperfusion. Amongst the symptoms are dizziness, headaches and possibly tachycardia. Orthostatic hypotension is often encountered in the older demographic.

Usually blood pressure is measured only once during routine checkups, punctually for extended or continuously for short periods of time – this is imposed by currently available diagnostic hardware. However, a continuous 24h history of blood pressure would yield superior results [Mann]. Enabling the reliable continuous trend estimation of blood pressure from a surface area comparable to an arterial line would improve the following aspects of vital signs monitoring: 1. Diminished risk and improved comfort for a considerable fraction of patients undergoing operative procedures; 2. Addition of blood pressure monitoring during medical procedures or everyday situations which do not justify the risks of using an arterial line, and 3. Obsolescence of the current generation of bulky continuous cuff-based devices requiring an air compressor.

How to achieve it?

Out blood pressure trend estimation is based on pulse wave transit time and morphological features picked up from singular, small footprint locations by means of multimodal, non-invasive data acquisition.

Primary pick-up methods include, but are not necessarily limited to reflective photoplethysmography (PPG), non-obtrusive mechanical pressure transduction and (multispectral) bioimpedance. Current anatomical regions of interest are the ulnar and radial artery at their superficial pass-through at the wrist; further usable locations are the carotid or the dorsalis pedis artery.

PPG is acquired monochromatically in a two dimensional array consisting of sensors and emitters, so as to obtain a “sufficiently” resolved image of the underlying arterial perfusion, both temporally and locally. Other than enabling a more detailed morphological analysis, this could yield the wave’s velocity and its morphological progression along the propagation path. A similar, yet redundancy-oriented setup, has already been evaluated.

Alternatively, or ideally complementary hereto, a similarly set-up array of miniature pressure transducers non-obtrusively palpates the pulsatility of the same artery. Multiple single transducers have been built and employed to successfully measure pulse wave velocity.

Combining longitudinal and transversal measurements of bioimpedance allows for an estimation of not only volumetric but also flow velocity and acceleration information of arterial blood. Single-frequency measurements have shown promising data in regard to peripheral pulse wave acquisition. Multispectral implementations are to follow.

Each of these methods is theoretically able to individually generate data with a certain degree of correlation to the arterial blood pressure. Other than adding an extra layer of redundancy to the system, using contextual information from all of the above-mentioned could improve the estimation quality.

As the processed signals are pulsatile in nature, data processing methods from the fields of autocorrelation, fractal analysis and wavelet transform are of major interest.

rBSN setup for ABP
rBSN setup for ABP

Related Work and References

[1] Pielmus, Alexandru-Gabriel, et al. "Novel Computation of Pulse Transit Time from Multi-Channel PPG Signals by Wavelet TransformCurrent Directions in Biomedical Engineering 2.1 (2016): 209-213.

[2] Kwok Leung Ong, et al. "Prevalence, Awareness, Treatment, and Control of Hypertension Among United States Adults 1999–2004" Hypertension (2007) 49: 69-75.

[3] Stewart Mann, et al. "Superiority of 24-hour Measurement of Blood Pressure Over Clinic Values in Determining Prognosis in Hypertension" Clinical and Experimental Hypertension. Part A: Theory and Practice, 7:2-3, 279-281

Zusatzinformationen / Extras


Schnellnavigation zur Seite über Nummerneingabe