Search

The present study aims to develop and validate a cuffless method for blood pressure continuous measurement through a wearable device. The goal is achieved according to the time-delay method, with the guiding principle of the time relation it takes for a blood volume to travel from the heart to a peripheral site. Inversely proportional to the blood pressure, this time relation is obtained as the time occurring between the R peak of the electrocardiographic signal and a marker point on the photoplethysmographic wave. Such physiological signals are recorded by using L.I.F.E. Italia’s wearable device, made of a sensorized shirt and wristband. A linear regression model is implemented to estimate the corresponding blood pressure variations from the obtained time-delay and other features of the photoplethysmographic wave. Then, according to the international standards, the model performance is assessed, comparing the estimates with the measurements provided by a certified digital sphygmomanometer. According to the standards, the results obtained during this study are notable, with 85% of the errors lower than 10 mmHg and a mean absolute error lower than 7 mmHg. In conclusion, this study suggests a time-delay method for continuous blood pressure estimates with good performance, compared with a reference device based on the oscillometric technique.

Hypertension is the leading cause of cardiovascular diseases in the world. To detect this disease easily and efficiently, it is necessary to obtain non invasive, continuous, automated and minimally intrusive blood pressure (BP) measurement. None of usual measurement methods succeed in validating all these criteria. Thus, the main goal of this PhD is to develop the proof-of-concept of a wearable device to measure BP. To do that, two different techniques are studied : photo-plethysmography (PPG), an optical technique enabling to obtain blood volume in arteries and ultrasound (US), an imaging technique currently used to obtain images of human tissue, like for example blood vessels. During my PhD, we succeed in showing that both modalities can be used to measure BP. For that, a multi-PPG device has been developed in CEA-Léti LS2P : first experiments with this device has confirmed its capacity to measure pulse wave velocity (PWV) and BP. Parallely, some tests have been done with US at CREATIS and enable to validate another method to measure PWV and BP. Finally, an experiment has been performed with a US probe and our multi-PPG device to test a multimodal processing that demonstrates the value of this bimodal BP measurement., L’hypertension artérielle (HTA) est la première cause de maladies cardiovasculaires au monde. Pour réussir à détecter cette maladie facilement et efficacement, il est primordial de réussir à mesurer la pression artérielle (PA) de manière non invasive, continue, automatisée et peu intrusive. Aucune des méthodes usuelles de mesure ne remplit tous ces critères. Ainsi, l’objectif principal de cette thèse est de développer la preuve-de-concept d’un dispositif connecté de mesure de PA. Pour ce faire, deux modalités sont étudiées : la photo-pléthysmographie (PPG), une technique optique donnant une image du volume de sang dans l’artère et les ultrasons (US), une technique d’imagerie couramment utilisée pour obtenir des images des tissus, par exemple les vaisseaux sanguins. Au cours de la thèse, il a été montré que les deux modalités investiguées peuvent être utilisées pour mesurer la PA. Pour ce faire, un dispositif multi-PPG a été développé au CEA Leti LS2P : des premières expériences avec le dispositif ont montré sa capacité à obtenir des valeurs de vitesse d’onde de pouls (VOP) et de PA. En parallèle, des tests ont été effectués en imagerie US ultra-rapide à CREATIS et ont permis de valider une autre méthode de mesure de VOP et de PA. Finalement, une expérience à partir d’une sonde US et de notre dispositif multi-PPG a permis de tester un traitement multimodal pour valider l’intérêt de cette mesure bimodale de PA.

The state-of-the-art blood pressure measurement is beside common cuff-based methods the cuff-less estimation of pulse-transit-time, which is the time a blood pressure wave requires to travel from left ventricle of the heart to another peripheral point in the cardiovascular system. Within this work we present a novel estimation method for cuff-less blood pressure measurement by analysing a single photoplethysmographic signal in the frequency domain. The harmonic phase-shift of the fundamental frequency and the first harmonic within the photoplethysmographic signal has proven a strong correlation of r = 0.8514 and r = 0.9315 with systolic and diastolic blood pressure respectively.

The purpose of this study was to evaluate the wrist-worn device Mio FUSE, which estimates heart rate (HR) based on photo-plethysmography, 1) in a large study group during a standardised activity, 2) in a small group during a variety of activities and 3) to investigate factors affecting HR accuracy in a real-world setting. First, 53 male participants (20 ±1 years; 1.79 ±0.07 m; 76.1 ±10.5 kg) completed a 35-km march wearing the Equivital EQ-02 as a criterion measure. Second, 5 participants (whereof 3 female; 29 ±5 years; 1.74 ±0.07 m; 67.8 ±11.1 kg) independently performed 25 activities, categorised as sitting passive, sitting active, standing, cyclic and anti-cyclic activities with the Polar H7 as a criterion device. Equivalence testing and Bland-and-Altman analyses were undertaken to assess the accuracy to the criterion devices. Third, confounders affecting HR accuracy were investigated using multiple backwards regression analyses. The Mio FUSE was equivalent to the respective criterion measures with only small systematic biases of -3.5 bpm (-2.6%) and -1.7 bpm (-1.3%) with limits of agreements of ±10.1 bpm and ±10.8 bpm during the 35-km march and during different activities, respectively. Confounding factors negatively affecting the accuracy of the Mio FUSE were found to include larger wrist size and intensified arm and/or wrist movement. The wrist-worn Mio FUSE can be recommended to estimate overall HR accurately for different types of activities in healthy adults. However, during sporting activities involving intensified arm and/or wrist movement or for detailed continuous analysis, a chest strap is preferred to the Mio FUSE to optimise HR estimation accuracy., Current Issues in Sport Science (CISS)