The diagnosis of hypertension and the adjustment of antihypertensive drugs are evolving from isolated measurements performed at the physician offices to the full phenotyping of patients in real-life conditions. Indeed, the strongest predictor of cardiovascular risk comes from night measurements. The aim of this study was to demonstrate that a wearable device (the Aktiia Bracelet) can accurately estimate BP in the most common body positions of daily life and thus become a candidate solution for the BP phenotyping of patients. We recruited 91 patients with BP ranging from low to hypertensive levels and compared BP values from the Aktiia Bracelet against auscultatory reference values for 4 weeks according to an extended ISO 81060-2 protocol. After initializing on day one, the observed means and standard deviations of differences for systolic BP were of 0.46 ± 7.75 mmHg in the sitting position, − 2.44 ± 10.15 mmHg in the lying, − 3.02 ± 6.10 mmHg in the sitting with the device on the lap, and − 0.62 ± 12.51 mmHg in the standing position. Differences for diastolic BP readings were respectively of 0.39 ± 6.86 mmHg, − 1.93 ± 7.65 mmHg, − 4.22 ± 6.56 mmHg and − 4.85 ± 9.11 mmHg. This study demonstrates that a wearable device can accurately estimate BP in the most common body positions compared to auscultation, although precision varies across positions. While wearable persistent BP monitors have the potential to facilitate the identification of individual BP phenotypes at scale, their prognostic value for cardiovascular events and its association with target organ damage will need cross-sectional and longitudinal studies. Deploying this technology at a community level may be also useful to drive public health interventions against the epidemy of hypertension.

The reduction of blood pressure is associated with a significant reduction of cardiovascular events and all-cause mortality. After lifestyle interventions, antihypertensive drugs are routinely administrated to lower blood pressure. The optimisation of antihypertensive therapies requires the identification of the best-adapted drugs for each patient, followed by the titration of doses or by the combination of drugs. Automated cuffless blood pressure monitoring devices based on optical sensors have the potential to provide new insights in the daily monitoring of the effectiveness of antihypertensive therapies over weeks. This case report illustrates the use of the CE-marked Aktiia Bracelet optical device for the monitoring of a 39-year-old male hypertensive patient for 4 months during which two drug therapies were tested.

Objective: Cuffless wrist-based blood pressure (BP) monitors have a great potential to make ambulatory monitoring more appealing to patients and provide clinicians with better BP profiling. Investigators must be vigilant to validate these devices properly, as little guidance has been published. We propose a protocol that addresses ambulatory environment challenges, such as body posture changes and variable arm position. In addition to reporting the accuracy, we suggest reporting the number of successful automated measurements – the acceptance rate. Design and method: We tested the Aktiia Bracelet, a cuffless automated device that records optical signals at the wrist and requires an initial calibration, with the proposed protocol on 10 healthy volunteers. We simultaneously recorded the signals from the Aktiia Bracelet and a reference volume-clamp device positioned on the contralateral arm during the following interventions: sitting, lying supine, standing with both arms positioned at heart level; sitting with the Aktiia bracelet wrist positioned 30 cm lower than heart level; isometric leg extension inducing large BP changes. We calculated the mean and standard deviations of the error between Aktiia bracelet and the reference, as well as the acceptance rates in different measurement scenarios. Results: The figure illustrates the mean and standard deviation (STD) of the error between Aktiia bracelet and the reference, as well as the acceptance rate (Acc) for the diastolic BP (DBP). The dashed area illustrates the expected hydrostatic bias. The mean and the standard deviation of the error fell within ISO81060-2 limits. Aktiia bracelet readings were not affected by the hydrostatic bias intervention, even though one could expect an error of more than 20 mmHg. The maximum of the accepted measurements was achieved when patients were lying supine (67%) and sitting (52%). Thus, the user can expect a high density of measurements during night-time. However, when standing, the acceptance dropped (27%). Conclusions: This protocol provides more realistic testing of cuffless wrist-based BP monitors and suggests a way to improve the validation of this new family of devices.

Objective The objective of this study (NCT04027777) was to assess the accuracy and precision of the Aktiia Bracelet, a CE-marked noninvasive optical blood pressure (BP) monitor worn at the wrist, over a period of 1 month. Methods In this study, participants aged between 21 and 65 years were recruited. The clinical investigation extended the ISO81060-2:2013 standard to the specificities of cuffless devices. Each BP assessment consisted of the simultaneous recording of optical signals with Aktiia Bracelet and double-blinded auscultation by two trained observers in the standard sitting position. The algorithms of Aktiia Bracelet further processed the recorded optical signals to perform a signal quality check and to calculate uncalibrated estimates of systolic BP (SBP) and diastolic BP (DBP). These estimates were transformed into mmHg using a subject-dependent calibration parameter, which was calculated using the first two available reference measurements per subject. Results Eighty-six participants were included in the analysis. The mean and SD of the differences between Aktiia Bracelet estimates and the reference (ISO81060-2 criterion 1) were 0.46 ± 7.75 mmHg for SBP and 0.39 ± 6.86 mmHg for DBP. The SD of the averaged paired difference per subject (ISO81060-2 criterion 2) were 3.9 mmHg for SBP and 3.6 mmHg for DBP. Conclusion After initialization and during 1 month, the overall accuracy of Aktiia Bracelet satisfied validation criteria 1 and 2 of ISO81060-2 in the sitting position. The Aktiia Bracelet can be recommended for BP measurement in the adult population.

Abstract Objective: The auscultatory blood pressure (BP) measurement technique, relying on the K1 and K5 Korotkoff sounds, is the current reference to validate new devices in the standard sitting position. There are few data to tell whether it should also be used for device validation in other body positions, since the K1 and K5 sounds can be affected by changes in vascular tone. In this study, we recorded the BP responses to body position changes by the auscultatory method, and we compared them to data from simultaneous recordings by a volume-clamp method. Design and method: Systolic (SBP) and diastolic (DBP) were estimated on the left upper arm by two independent blinded observers, and on the ipsilateral middle finger by a volume-clamp device (Nexfin, BMEYE, The Netherlands) in the supine followed by the standing position, and in the standing followed by the supine position. The auscultatory readings were repeated if readings differed more than 4mmHg between the two observers. A first auscultation was performed before the position change, and a second one 150 s after the position change. Because volume-clamp measurements were not available during upper-arm cuff inflation, the mean of volume-clamp values in a 30 s window prior to auscultation onset were used for the analyses. Results: Sixty-seven participants, between 21 and 65 years old, participated in this study. As illustrated in the figure, volume-clamp reported a consistent mean DBP increase of 9.3 mmHg when moving from supine to sitting, and a consistent mean DBP decrease of 10.4 mmHg when moving from sitting to supine. These results agree with those previously reported using simultaneous invasive and volume-clamp measurements. On the contrary, auscultation detected no consistent body position-related change in the DBP readings. Neither auscultation nor volume-clamp detected consistent body position-related changes in SBP readings. Conclusions: Compared to volume-clamp, auscultation was not able to detect any consistent changes in BP during orthostatic challenge. Our study suggests that the use of Korotkoff sounds to estimate BP in body positions other than sitting and relaxed may not be appropriate.

Objective The objective of this study was to compare the systolic (S) and diastolic (D) blood pressure (BP) estimations from a new optical device at the wrist with invasive measurements performed on patients scheduled for radial arterial catheterization in the ICU. Optical signals were automatically processed by a library of algorithms from Aktiia SA (OBPM – optical blood pressure monitoring algorithms). Methods A total of 31 participants from both sexes, aged 32–87 years, were enrolled in the study (NCT03837769). The measurement protocol consisted of the simultaneous recording of reflective photoplethysmographic signals (PPG) from the cuffless optical device and the reference BP values recorded by a contralateral radial arterial catheter. From the 31 participants, 23 subjects whose reference data quality requirements were adequate were retained for further analysis. The PPG signals from these patients were then automatically processed by the Aktiia OBPM library of algorithms, which generated uncalibrated estimates of SBP and DBP. After the automatic assessment of optical signal quality, 326 pairs of uncalibrated SBP and DBP determinations from 16 patients were available for analysis. These values were finally transformed into calibrated estimations (in mmHg) using arterial catheter SBP and DBP values, respectively. Results For SBP, a mean difference (±SD) of 0.0 ± 7.1 mmHg between the arterial catheter and the optical device values was found, with 95% limits of agreement in the Bland-Altman method of –11.9 to + 12.2 mmHg (correlation of r = 0.87, P < 0.001). For DBP, a mean difference (±SD) of 0.0 ± 2.9 mmHg between arterial catheter and the optical device values was found, with 95% limits of agreement in the Bland-Altman method of –4.8 to + 5.5 mmHg (correlation of r = 0.98, P < 0.001). Conclusion SBP and DBP values obtained by radial artery catheterization and those obtained from optical measurements at the wrist were compared. The new optical technique appears to be capable of replacing more traditional methods of BP estimation.

Double-blinded auscultation is the current reference to validate new devices in the sitting position. There are few data to tell whether it should be used for device validation in other body positions, as the Korotkoff sounds can be affected by changes in vascular tone. In this study, we recorded the BP response to orthostatic posture change (standing to supine) in 75 subjects, aged between 21 and 65 years old. Systolic (SBP) and diastolic (DBP) were measured on the left upper arm by auscultation by two independent blinded observers before and 150s after posture change. In case the observers did not agree, i.e. readings differed more than 4 mmHg, the measurement was repeated. Beat-to-beat BP values were measured on the ipsilateral middle finger with Nexfin (BMEYE, The Netherlands). Because Nexfin measurements were not available during upper-arm cuff inflation, the mean of the values in a 30s window prior to auscultation onset was used for the analysis. The distribution of the BP responses to posture change was characterized in terms of median, 10 th and 90 th percentiles (see Table). The response was considered consistent if these percentiles were on the same side of zero. Neither auscultation nor Nexfin detected any consistent posture-related changes in SBP when going from standing to supine. Nexfin detected a consistent decrease in DBP. Auscultation detected no consistent posture-related change. Compared to volume-clamp, auscultation was not able to detect any consistent changes in DBP during orthostatic challenge. Our study suggests that the use of Korotkoff sounds to estimate BP in body positions other than sitting may not be appropriate.

This study aims at evaluating the potential of a wrist-type photoplethysmographic (PPG) device to discriminate between atrial fibrillation (AF) and other types of rhythm. Data from 17 patients undergoing catheter ablation of various arrhythmias were processed. ECGs were used as ground truth and annotated for the following types of rhythm: sinus rhythm (SR), AF, and ventricular arrhythmias (VA). A total of 381/1370/415 10-s epochs were obtained for the three categories, respectively. After pre-processing and removal of segments corresponding to motion artifacts, two different types of feature were derived from the PPG signals: the interbeat interval-based features and the wave-based features, consisting of complexity/organization measures that were computed either from the PPG waveform itself or from its power spectral density. Decision trees were used to assess the discriminative capacity of the proposed features. Three classification schemes were investigated: AF against SR, AF against VA, and AF against (SR&VA). The best results were achieved by combining all features. Accuracies of 98.1/95.9/95.0 %, specificities of 92.4/88.7/92.8 %, and sensitivities of 99.7/98.1/96.2 % were obtained for the three aforementioned classification schemes, respectively.

In this paper, we propose a fast novel nonlinear filtering method named Relative-Energy (Rel-En), for robust short-term event extraction from biomedical signals. We developed an algorithm that extracts short- and long-term energies in a signal and provides a coefficient vector with which the signal is multiplied, heightening events of interest. This algorithm is thoroughly assessed on benchmark datasets in three different biomedical applications, namely ECG QRS-complex detection, EEG K-complex detection, and imaging photoplethysmography (iPPG) peak detection. Rel-En successfully identified the events in these settings. Compared to the state-of-the-art, better or comparable results were obtained on QRS-complex and K-complex detection. For iPPG peak detection, the proposed method was used as a preprocessing step to a fixed threshold algorithm that lead to a significant improvement in overall results. While easily defined and computed, Rel-En robustly extracted short-term events of interest. The proposed algorithm can be implemented by two filters and its parameters can be selected easily and intuitively. Furthermore, Rel-En algorithm can be used in other biomedical signal processing applications where a need of short-term event extraction is present.

Imaging photoplethysmography (iPPG) is a promising technology for contactless heart rate (HR) monitoring. However iPPG signals are easily deteriorated by subject movements and illumination changes. The purpose of this study was to develop a signal quality index (SQI) for realtime HR monitoring applications and to assess its performance on a challenging dataset composed of videos of moving subjects. HR was estimated using a multi-input adaptive frequency tracking scheme, in which the iPPG signals derived with different methods and their corresponding SQIs were provided as inputs. Using the proposed SQI, the average absolute error was reduced by 42%/45% when the forehead/entire face region was used to derive iPPG signals, respectively.

Photoplethysmographic (PPG) signals are easily corrupted by motion artifacts when the subjects perform physical exercise. This paper introduces a two-step processing scheme to estimate heart rate (HR) from wrist-type PPG signals strongly corrupted by motion artifacts. Adaptive noise cancellation, using normalized least-mean-square algorithm, is first performed to attenuate motion artifacts and reconstruct multiple PPG waveforms from different combinations of corrupted PPG waveforms and accelerometer data. An adaptive band-pass filter is then used to track the common instantaneous frequency component (i.e. HR) of the reconstructed PPG waveforms. The proposed HR estimation scheme was evaluated on two datasets, composed of records from running subjects and subjects performing different kinds of arm/forearm movements and resulted in average absolute errors of 1.40  ±  0.60 and 4.28  ±  3.16 beats-per-minute for these two datasets, respectively. Importantly, the proposed method is fully automatic, induces an average estimation delay of 0.93 s, and is therefore suitable for real-time monitoring applications.

Imaging photoplethysmography (iPPG) has gained a lot of popularity as a contactless heart rate (HR) monitoring technique. However, most of the existing approaches to estimate HR are based on block-wise processing schemes, which are not optimal for real-time applications. The aim of this study was to investigate robust HR estimation methods having a short estimation delay, which would be suitable for real-time HR monitoring applications using iPPG. The three following algorithms were evaluated: 1) an algorithm based on adaptive sliding-window singular value decomposition (SWASVD), 2) an adaptive band-pass filter (OSC-ANF-W), 3) a notch-filter bank (NFB) estimation method. The database used to evaluate these algorithms was composed of 46 records, acquired in the light using an RGB camera or the dark using an NIR camera. The subjects were asked to perform different tasks to induce HR fluctuations. For the visible/dark sequences, average absolute errors (AAE) of 3.42/5.25, 3.14/4.21 and 3.98/6.02 bpm were obtained for the SWASVD, the OSC-ANF-W and the NFB algorithms, respectively. The corresponding averaged estimation delays were 4 seconds for the SWASVD and OSC-ANF-W, and 3 seconds for the NFB.

The purpose of this study was to develop algorithms to lower the incidence of false arrhythmia alarms in the ICU using information from independent sources, namely electrocardiogram (ECG), arterial blood pressure (ABP) and photoplethysmogram (PPG). Our approach relies on robust adaptive signal processing techniques in order to extract accurate heart rate (HR) values from the different waveforms. Based on the quality of available signals, heart rate was either estimated from pulsatile waveforms using an adaptive frequency tracking algorithm or computed from ECGs using an adaptive mathematical morphology approach. Furthermore, we developed a supplementary measure based on the spectral purity of the ECGs to determine whether a ventricular tachycardia or flutter/fibrillation arrhythmia has taken place. Finally, alarm veracity was determined based on a set of decision rules on HR and spectral purity values. The proposed method was evaluated on the PhysioNet/CinC Challenge 2015 database, which is composed of 1250 life-threatening alarm recordings, each categorized into either bradycardia, tachycardia, asystole, ventricular tachycardia or ventricular flutter/fibrillation arrhythmia. This resulted in overall true positive rates of 95%/99% and overall true negative rates of 76%/80% on the real-time and retrospective subsets of the test dataset, respectively.

Imaging photoplethysmography (iPPG) has emerged as a contactless heart-rate monitoring technique. As the respiratory activity modulates the heart rate, we investigate the accuracy of iPPG in conveying the inter-beat variation due to the respiratory modulation of the heart rate. The instantaneous respiratory rate was estimated in real-time from the iPPG inter-beat variations with an algorithm based on a bank of short FIR notch filters. The comparison of the iPPG-based respiratory rate estimates to ECG-based estimates showed that the iPPG ones were only slightly less accurate in spite of the challenging conditions related to this contacless technique.

The present study aimed to compare sleep disordered breathing during live high-train low (LHTL) altitude camp using normobaric hypoxia (NH) and hypobaric hypoxia (HH). Sixteen highly trained triathletes completed two 18-day LHTL camps in a crossover designed study. They trained at 1100-1200 m while they slept either in NH at a simulated altitude of 2250 m or in HH. Breathing frequency and oxygen saturation (SpO(2)) were recorded continuously during all nights and oxygen desaturation index (ODI 3%) calculated. Breathing frequency was lower for NH than HH during the camps (14.6 +/- 3.1 breath x min(-1) vs. 17.2 +/- 3.4 breath x min(-1), p < 0.001). SpO(2) was lower for HH than NH (90.8 +/- 0.3 vs. 91.9 +/- 0.2, p < 0.001) and ODI 3% was higher for HH than NH (15.1 +/- 3.5 vs. 9.9 +/- 1.6, p < 0.001). Sleep in moderate HH is more altered than in NH during a LHTL camp.

The potential of imaging photoplethysmography for cardiovascular monitoring applications has been demonstrated recently. Various processing schemes have been proposed to extract heart rate (HR) from a defined region of interest (ROI) on the face. However, the reasons that motivate the choice of the ROI are often unclear. This study aimed at investigating the spatial distribution of the HR-related information on the subject face, based on a partition into 260 small ROIs. A power spectral density (PSD) analysis was performed to determine the amount of HR-related information in each ROI. Normalized color maps were used to visualize the spatial distribution of the HR-information and clearly showed that color fluctuations due to blood volume changes are always more pronounced in the forehead region. After face segmentation, for the R/G/B/NIR channels, average percentages of power of 27%b/43%b/27%b/36%b for the forehead region, 17%/28%/18%/21%for the cheek region and 16%/24%/16%/20% for the whole face were obtained.

This study aims at evaluating the performances of a wrist-located device to detect atrial fibrillation (AF) based on photoplethysmography (PPG) technology. Twenty patients referred for catheter ablation of cardiac arrhythmias in whom episodes of sinus rhythm (SR) and AF coexisted were screened. Screening included a 12-lead electrocardiogram (ECG) and a PPG device placed at the wrist measuring cardiac pulsatility by means of infrared light. While reference cardiac interbeat (RR) intervals were obtained from the analysis of the ECG signals, RR intervals from the PPG signals were estimated by detecting systolic down-strokes on the optical waveforms. Classification of SR versus AF epochs was obtained via a support vector machine to which features extracted on 10-second windows were provided. Extracted features included mean, standard deviation, minimum, and interquartile range of RR within an epoch. A total number of 2213 epochs (1927 of AF, 286 of SR) were analyzed, providing classification accuracy of 93.85% for the PPG-based classifier and 98.93% for the ECG-based classifier. These preliminary results suggest that a wrist-located PPG-based monitor might be eligible for future screening of AF in large populations.

The performance of photoplethysmography (PPG)-based wearable monitors to diagnose atrial fibrillation (AF) remains unknown to date. This study aims at assessing the performance of new indices quantifying the level of organization in PPG signals to diagnose AF. A database made of 18 adult patients undergoing catheter ablation of various cardiac arrhythmias was used. PPG signals were recorded using a wrist-type sensor A 12-lead ECG was used as gold standard. ECGs were annotated by experts and selected segments were divided into 4 categories: sinus rhythm (SR), regularly paced rhythm (RPR), irregularly paced rhythm (IPR) and AF. The level of organization of the various PPG signals was measured using an adaptive organization index (AOI), defined as the ratio of the power of the fundamental frequency and the first harmonic to the total power of the PPG signal, computed with adaptive band-pass filters. A total of 2806/803/852/287 10-second epochs were considered for AF/SR/RPR/IPR classes. The following mean AOI values were measured: 0.45±0.11 for AF, 0.73±0.19 for SR, 0.78±0.20 for RPR and 0.61±0.19 for IPR classes. Importantly, the AF AOI was significantly smaller than that of the other categories (p<0.001), indicating a higher degree of disorganization.

As part of the 2015 PhysioNet/CinC Challenge, this work aims at lowering the number of false alarms, which are a persistent concern in the intensive care unit. The multimodal database consists of 1250 life-threatening alarm recordings, each categorized as a bradycardia, tachycardia, asystole, ventricular tachycardia or ventricular flutter/fibrillation arrhythmia. Based on the quality of available signals, heart rate was either estimated from pulsatile waveforms (photoplethysmogram and/or arterial blood pressure) using an adaptive frequency tracking algorithm or computed from ECGs using an adaptive mathematical morphology approach. Furthermore, we introduced a supplementary measure based on the spectral purity of the ECGs to determine if a ventricular tachycardia or flutter/fibrillation arrhythmia has taken place. Finally, alarm veracity was determined based on a set of decision rules on heart rate and spectral purity values. Our method achieved overall scores of 76.11 and 85.04 on the real-time and retrospective subsets, respectively.