Wearable estimation of central aortic blood pressure (Estimación vestible de la presión arterial central aórtica)
German Fierro
PhD thesis from Universidad de la República (Uruguay). Facultad de Ingeniería. IIE
Advisor(s): Fernando Silveira, Ricardo L. Armentano
- Mar. 2020
Research group(s):  Microelectronica (gme)
Department(s):  Electrónica
Download the publication : Fie20.pdf [10.1MB]  

Abstract

Arterial hypertension affects a third of the world's population and is a significant risk factor for cardiovascular disease. Blood pressure (BP) is one of the most relevant parameters used for monitoring of possible hypertension states in patients at risk of cardiovascular disease. Hence, there exists a need for new monitoring solutions, which allow to increase the frequency between BP assessments, but also allow to reduce the level of occlusion in the attempts. Moens-Korteweg equation is among the main principles to estimate BP by dispensing of any inflatable cuff. This principle might lead to an indirect estimation of BP by measuring the time it takes the pressure pulse to propagate between two pre-established vascular points, accordingly the pulse transit time (PTT) method. This thesis proposes a wearable PTT-based method to estimate central aortic BP (CABP) and, the main milestones of this work included: proof of concept of the proposed method (pilot work), the development of a wearable device (including two stages of validation), the proposition of a miniaturized version (integrated circuit) of the analog front-end of the wearable hardware, and, the development of a novel PTT-based model (PTTBM, i.e., the mathematical relationship between measured variables and estimated BP) suitable for the proposed wearable methodology to estimate BP. The main contributions found at each milestone are presented. One of the contributions of this thesis is the use of the PTT-principle for estimating CABP instead of the peripheral BP (PBP) (as typically used in the literature). The pilot work showed the feasibility of CABP estimation from the PTT principle by using electrocardiogram (ECG) and ballistocardiogram (BCG) recordings from off-the-shelf equipment. Results showed that CABP was more correlated with the proposed methodology in comparison to all PBP variables assessed; confirming our hypothesis that the CABP is the most suitable parameter to collate through the time elapsed from ECG R-wave to the BCG J-wave. That is, considered featured time (RJ-interval) includes the time of a pulse pressure propagating at an aortic district. Bland-Altman plots showed an almost zero mean error (\u\ < 0.02mmHg) and bounded standard deviation o < 5mmHg for all systolic and mean central BP readings. Pilot work provided a landmark in order to develop a compact device that allows the integration of wireless blood pressure monitoring into a wearable system. Another contribution of this thesis is the proposition of a wearable device for PTT-computing by also including design considerations for the signal conditioning chains for ECG and BCG signals. The proposed design procedure takes care of minimizing the impact of spurious delays between physiological signals, which eventually degrade the PTT computation. Further, such a procedure could be suitable for any PTT-acquisition. Filtering with low and controlled delay is required for this biomedical application, and proposed conditioning chains provide less than 2ms group-delay, showing the effectiveness of the proposed approach. In order to provide the methodology with higher autonomy and integration, a highly miniaturized implementation of the filtering approach was also proposed. It includes the design of proposed architectures in CMOS technology to implement the particular low-delay filtering at reduced bandwidth featuring ultra-low-power characteristics. Results show that less than 2ms delay for the ECG QRS-complex can be achieved with a total current consumption of IDD = 2:1nA at VDD = 1:2V of power supply. Such development meant another significant contribution of this work in the conception of highly autonomous wearable devices for PTT acquisition. The first stage of validations on the wearable CABP estimation showed that, when considering data from one volunteer, results achieved with off-the-shelf equipment could be replicated by using a proposed wearable device, and the method could be further validated by using the wearable version. Additionally, CABP estimation from the proposed wearable device could be feasible by using three feature times (FTs) as CABP surrogates; that is, RI, RJ, and IJ intervals (from ECG and BCG wearable recordings). The first validation of the method also showed that CABP could be accurately predicted by the proposed methodology when in the order of daily calibrations are performed. The second stage of validations involved a study with a group of volunteers, and new alternatives were explored (twentyseven: nine PTTBMs along the three FTs) for the CABP estimation. We found that CABP could be accurately estimated (inside AAMI requirements) through the presented methodology by using four of the explored alternatives, whereas the RI interval, an FT lacking any PTT assessment, emerged as the best surrogate for the CABP estimation. Hence, a principle different from the traditional PTT-based method arises as a more advantageous method for the CABP estimation in the light of evidence reported in this validation, and, to our knowledge, this is the first time that CABP has been successfully estimated from a wearable device. The final significant contribution of this thesis meant the last chain-link in the process to achieve an utterly original method to estimate CABP. A novel PTTBM to estimate CABP is proposed, which uses a ow-driven two-element Windkesel network constructed from FTs extracted from the wearable recordings. When classic PTTBMs are applied, the fitting of parameters often leads to values without a physiological basis. Opposite to that in the proposed PTTBM, the parameters have a clear physiological meaning, and the parameter fitting led to values that are consistent with this meaning and more stable throughout calibrations. In conclusion, this thesis introduces a novel device that exploits an alternative and indirect method for CABP estimation. Variants of the principle used, accordingly, PTT method, have been previously explored to estimate PBP but not for central aortic BP. Additionally, the device was designed to be wearable; that is, it is attached to the clothes, causing low discomfort for the user during the measurement, thus, allowing continuous and ambulatory monitoring of aortic pressure. The developed wearable system, validated in a series of volunteers, showed promising results towards the continuous CABP monitoring.

Additional data

""

BibTex references

Descargar BibTex bibtex

Other publications in the database

» German Fierro