Assessing hemodynamics from the photoplethysmogram to gain insights into vascular age: a review from VascAgeNet

Charlton P.H., Marozas V.

The wearables market has expanded greatly in recent years, with wrist-worn devices now widely used. Smart wearables provide opportunity to monitor health and fitness in daily life. Wearables such as fitness bands and smartwatches routinely monitor the photoplethysmogram (PPG) signal, an optical measure of the arterial pulse wave which is strongly influenced by the heart and blood vessels. This chapter presents a comprehensive overview of the state-of-the-art of wearable photoplethysmography devices. It summarizes: (i) key considerations in the design of wearable PPG devices; (ii) the physiological parameters that can be estimated from wearable PPG signals; (iii) commercially available devices; and (iv) potential applications in health and fitness monitoring.

Kamboj N., Chang K., Metcalfe K., Chu C.H., Conway A.

To summarize the evidence regarding the accuracy of continuous non-invasive arterial pressure measurements in adult critical care patients.Medline, EMBASE, and CINAHL were searched for studies that included adult critical care patients reporting the agreement between continuous non-invasive and invasive arterial pressure measurements. The studies were selected and assessed for risk of bias using the Revised Quality Assessment of Diagnostic Accuracy Studies tool by two independent reviewers. The Grading of Recommendations, Assessment, Development and Evaluations approach was used. Pooled estimates of the mean bias and limits of agreement with outer 95% confidence intervals (termed population limits of agreement) were calculated.Population limits of agreement for systolic blood pressure were wide, spanning from -36.13 mmHg to 28.28 mmHg (18 studies; 785 participants). Accuracy of diastolic blood pressure measurements was highly inconsistent across studies, resulting in imprecise estimates for the population limits of agreement. Population limits of agreement for mean arterial pressure spanned from -39.96 mmHg to 44.36 mmHg (17 studies; 765 participants). The evidence was rated as very low-quality due to very serious concerns about heterogeneity and imprecision.Substantial differences in blood pressure were identified between measurements taken from continuous non-invasive and invasive monitoring devices. Clinicians should consider this broad range of uncertainty if using these devices to inform clinical decision-making in critical care.

Miyasaka K., Shelley K., Takahashi S., Kubota H., Ito K., Yoshiya I., Yamanishi A., Cooper J.B., Steward D.J., Nishida H., Kiani J., Ogino H., Sata Y., Kopotic R.J., Jenkin K., et. al.

Dr. Takuo Aoyagi invented pulse oximetry in 1974. Pulse oximeters are widely used worldwide, most recently making headlines during the COVID-19 pandemic. Dr. Aoyagi passed away on April 18, 2020, aware of the significance of his invention, but still actively searching for the theory that would take his invention to new heights. Many people who knew Dr. Aoyagi, or knew of him and his invention, agreed to participate in this tribute to his work. The authors, from Japan and around the world, represent all aspects of the development of medical devices, including scientists and engineers, clinicians, academics, business people, and clinical practitioners. While the idea of pulse oximetry originated in Japan, device development lagged in Japan due to a lack of business, clinical, and academic interest. Awareness of the importance of anesthesia safety in the US, due to academic foresight and media attention, in combination with excellence in technological innovation, led to widespread use of pulse oximetry around the world. Dr. Aoyagi’s final wish was to find a theory of pulse oximetry. We hope this tribute to him and his invention will inspire a new generation of scientists, clinicians, and related organizations to secure the foundation of the theory.

Kontaxis S., Gil E., Marozas V., Lazaro J., Garcia E., Posadas-de Miguel M., Siddi S., Bernal M.L., Aguilo J., Haro J.M., de la Camara C., Laguna P., Bailon R.

Objective: In the present study, a photoplethysmographic (PPG) waveform analysis for assessing differences in autonomic reactivity to mental stress between patients with Major Depressive Disorder (MDD) and healthy control (HC) subjects is presented. Methods: PPG recordings of 40 MDD and 40 HC subjects were acquired at basal conditions, during the execution of cognitive tasks, and at the post-task relaxation period. PPG pulses are decomposed into three waves (a main wave and two reflected waves) using a pulse decomposition analysis. Pulse waveform characteristics such as the time delay between the position of the main wave and reflected waves, the percentage of amplitude loss in the reflected waves, and the heart rate (HR) are calculated among others. The intra-subject difference of a feature value between two conditions is used as an index of autonomic reactivity. Results: Statistically significant individual differences from stress to recovery were found for HR and the percentage of amplitude loss in the second reflected wave ( $A_{13}$ ) in both HC and MDD group. However, autonomic reactivity indices related to  $A_{13}$ reached higher values in HC than in MDD subjects (Cohen's  ${d=\!-0.81},{\,\text {AUC}=0.74}$ ), implying that the stress response in depressed patients is reduced. A statistically significant ( $p \,<\, 0.001$ ) negative correlation ( $r=-0.5$ ) between depression severity scores and $A_{13}$ was found. Conclusion: A decreased autonomic reactivity is associated with higher degree of depression. Significance: Stress response quantification by dynamic changes in PPG waveform morphology can be an aid for the diagnosis and monitoring of depression.

Huthart S., Elgendi M., Zheng D., Stansby G., Allen J.

Objective: Despite the vast number of photoplethysmography (PPG) research publications and growing demands for such sensing in Digital and Wearable Health platforms, there appears little published on signal quality expectations for morphological pulse analysis. Aim: to determine a consensus regarding the minimum number of undistorted i.e., diagnostic quality pulses required, as well as a threshold proportion of noisy beats for recording rejection.Approach: Questionnaire distributed to international fellow researchers in skin contact PPG measurements on signal quality expectations and associated factors concerning recording length, expected artifact-free pulses (“diagnostic quality”) in a trace, proportion of trace having artifact to justify excluding/repeating measurements, minimum beats required, and number of respiratory cycles.Main Results: 18 (of 26) PPG researchers responded. Modal range estimates considered a 2-min recording time as target for morphological analysis. Respondents expected a recording to have 86–95% of diagnostic quality pulses, at least 11–20 sequential pulses of diagnostic quality and advocated a 26–50% noise threshold for recording rejection. There were broader responses found for the required number of undistorted beats (although a modal range of 51–60 beats for both finger and toe sites was indicated).Significance: For morphological PPG pulse wave analysis recording acceptability was indicated if &lt;50% of beats have artifact and preferably that a minimum of 50 non-distorted PPG pulses are present (with at least 11–20 sequential) to be of diagnostic quality. Estimates from this knowledge transfer exercise should help inform students and researchers as a guide in standards development for PPG study design.

Natarajan A., Pantelopoulos A., Emir-Farinas H., Natarajan P.

Summary Background Heart rate variability, or the variation in the time interval between consecutive heart beats, is a non-invasive dynamic metric of the autonomic nervous system and an independent risk factor for cardiovascular death. Consumer wrist-worn tracking devices using photoplethysmography, such as Fitbit, now provide the unique potential of continuously measuring surrogates of sympathetic and parasympathetic nervous system activity through the analysis of interbeat intervals. We aimed to leverage wrist-worn trackers to derive and describe diverse measures of cardiac autonomic function among Fitbit device users. Methods In this cross-sectional study, we collected interbeat interval data that are sent to a central database from Fitbit devices during a randomly selected 24 h period. Age, sex, body-mass index, and steps per day in the 90 days preceding the measurement were extracted. Interbeat interval data were cleaned and heart rate variability features were computed. We analysed heart rate variability metrics across the time (measured via the root mean square of successive RR interval differences [RMSSD] and SD of the RR interval [SDRR]), frequency (measured by high-frequency and low-frequency power), and graphical (measured by Poincare plots) domains. We considered 5 min windows for the time and frequency domain metrics and 60 min measurements for graphical domain metrics. Data from participants were analysed to establish the correlation between heart rate variability metrics and age, sex, time of day, and physical activity. We also determined benchmarks for heart rate variability (HRV) metrics among the users. Findings We included data from 8 203 261 Fitbit users, collected on Sept 1, 2018. HRV metrics decrease with age, and parasympathetic function declines faster than sympathetic function. We observe a strong diurnal variation in the heart rate variability. SDRR, low-frequency power, and Poincare S2 show a significant variation with sex, whereas such a difference is not seen with RMSSD, high-frequency power, and Poincare S1. For males, when measured from 0600 h to 0700 h, the mean low-frequency power decreased by a factor of 66·5% and high-frequency power decreased by a factor of 82·0% from the age of 20 years to 60 years. For females, the equivalent factors were 69·3% and 80·9%, respectively. Comparing low-frequency power between males and females at the ages of 40–41 years, measured from 0600 h to 0700 h, we found excess power in males, with a Cohen's d effect size of 0·33. For high-frequency power, the equivalent effect size was −0·04. Increased daily physical activity, across age and sex, was highly correlated with improvement in diverse measures of heart rate variability in a dose-dependent manner. We provide benchmark tables for RMSSD, SDRR, high and low frequency powers, and Poincare S1 and S2, separately for different ages and sex and computed at two times of the day. Interpretation Diverse metrics of cardiac autonomic health can be derived from wrist-worn trackers. Empirical distributions of heart rate variability can potentially be used as a framework for individual-level interpretation. Increased physical activity might yield improvement in heart rate variability and requires prospective trials for confirmation. Funding Fitbit.

Mishra T., Wang M., Metwally A.A., Bogu G.K., Brooks A.W., Bahmani A., Alavi A., Celli A., Higgs E., Dagan-Rosenfeld O., Fay B., Kirkpatrick S., Kellogg R., Gibson M., Wang T., et. al.

Consumer wearable devices that continuously measure vital signs have been used to monitor the onset of infectious disease. Here, we show that data from consumer smartwatches can be used for the pre-symptomatic detection of coronavirus disease 2019 (COVID-19). We analysed physiological and activity data from 32 individuals infected with COVID-19, identified from a cohort of nearly 5,300 participants, and found that 26 of them (81%) had alterations in their heart rate, number of daily steps or time asleep. Of the 25 cases of COVID-19 with detected physiological alterations for which we had symptom information, 22 were detected before (or at) symptom onset, with four cases detected at least nine days earlier. Using retrospective smartwatch data, we show that 63% of the COVID-19 cases could have been detected before symptom onset in real time via a two-tiered warning system based on the occurrence of extreme elevations in resting heart rate relative to the individual baseline. Our findings suggest that activity tracking and health monitoring via consumer wearable devices may be used for the large-scale, real-time detection of respiratory infections, often pre-symptomatically. Analysis of physiological and activity data from consumer smartwatches enables real-time detection, often before symptom onset, of COVID-19, as well as other respiratory illnesses and stress inducers.

Charlton P.H., Kyriacou P., Mant J., Alastruey J.

Kulin D., Antali F., Kulin S., Wafa D., Lucz K.I., Veres D.S., Miklós Z.

Telemonitoring systems equipped with photoplethysmography-based contour analysis of the digital arterial volume pulse (DVP) can be optimal tools for remote monitoring of cardiovascular patients; however, the method is known to be sensitive to errors. We aimed to show that DVP analysis is a reliable method to track cardiovascular status. We used our proprietary SCN4ALL telemedicine system and analyzed nine parameters derived from the DVP and its second derivative (SDDVP). First, we assessed the repeatability of system measurements by detecting artificial signals. Then test–retest reliability of human measurements was evaluated in healthy individuals under standardized conditions. The SCN4ALL system analyzed each parameter with high accuracy (coefficients of variation (CVs) < 1%). Test–retest reliability of most parameters (stiffness index, reflection index, left ventricular ejection time index, b/a, heart rate) was satisfactory (CVs < 10%) in healthy individuals. However, aging index and d/a ratio derived from the SDDVP were more variable. Photoplethysmography-based pulse contour analysis is a reliable method to monitor cardiovascular status if measurements are performed with a system of high accuracy. Our results highlighted that SDDVP parameters can be interpreted with limitations due to (patho)physiological variations of the DVP. We recommend the evaluation of these parameters only in measurements where all inflections of SDDVP are detected reliably.

Chandrasekhar A., Yavarimanesh M., Natarajan K., Hahn J., Mukkamala R.

Objective: Photo-plethysmography (PPG) sensors are often used to detect pulse transit time (PTT) for potential cuff-less blood pressure (BP) measurement. It is known that the contact pressure (CP) of the PPG sensor markedly alters the PPG waveform amplitude. The objective was to test the hypothesis that PTT detected via PPG sensors is likewise impacted by CP. Methods: A device was built to measure the time delay between ECG and finger PPG waveforms (i.e., pulse arrival time (PAT) - a popular surrogate of PTT) and the PPG sensor CP at different CP levels. These measurements and finger cuff BP were recorded while the CP was slowly varied in 17 healthy subjects. Results: Over a physiologic range of CP, the maximum deviations of PAT detected at the PPG foot and peak were 22±2 and 40±7 ms (p

Quer G., Radin J.M., Gadaleta M., Baca-Motes K., Ariniello L., Ramos E., Kheterpal V., Topol E.J., Steinhubl S.R.

Traditional screening for COVID-19 typically includes survey questions about symptoms and travel history, as well as temperature measurements. Here, we explore whether personal sensor data collected over time may help identify subtle changes indicating an infection, such as in patients with COVID-19. We have developed a smartphone app that collects smartwatch and activity tracker data, as well as self-reported symptoms and diagnostic testing results, from individuals in the United States, and have assessed whether symptom and sensor data can differentiate COVID-19 positive versus negative cases in symptomatic individuals. We enrolled 30,529 participants between 25 March and 7 June 2020, of whom 3,811 reported symptoms. Of these symptomatic individuals, 54 reported testing positive and 279 negative for COVID-19. We found that a combination of symptom and sensor data resulted in an area under the curve (AUC) of 0.80 (interquartile range (IQR): 0.73–0.86) for discriminating between symptomatic individuals who were positive or negative for COVID-19, a performance that is significantly better (P < 0.01) than a model1 that considers symptoms alone (AUC = 0.71; IQR: 0.63–0.79). Such continuous, passively captured data may be complementary to virus testing, which is generally a one-off or infrequent sampling assay. A smartphone app that combines smartwatch and activity tracker data together with self-reported symptoms allows continuous monitoring of SARS-CoV-2 infection.

Bruno R.M., Nilsson P.M., Engström G., Wadström B.N., Empana J., Boutouyrie P., Laurent S.

Pulse wave velocity is an established marker of early vascular aging but may also help identifying individuals with supernormal vascular aging. We tested the hypothesis that individuals with the largest difference (Δ-age) between chronological and vascular age show the lowest rate of cardiovascular events and may thus be defined as supernormal vascular aging. Vascular age was defined as the predicted age in the best fitting multivariable regression model including classical risk factors and treatment and pulse wave velocity, in a subset of the Reference Values for Arterial Stiffness Collaboration Database (n=3347). Δ-age was then calculated as chronological age minus vascular age, and the 10th and 90th percentiles were used to define early (Δ-age<−5.7 years), normal (Δ-age −5.7 to 6.8 years) and supernormal vascular aging (Δ-age>6.8 years). The risk for fatal and nonfatal cardiovascular events associated with vascular aging categories was investigated in the Malmö Diet and Cancer Study cohort (n=2642). In the Malmö Diet and Cancer Study Cohort (6.6-year follow-up, 286 events), Δ-age was significantly ( P <0.01) and inversely associated with cardiovascular events. Compared with normal vascular aging, supernormal vascular aging had lower risk (hazard ratio, 0.59 [95% CI, 0.41–0.85]), whereas early vascular aging had higher risk (hazard ratio, 2.70 [95% CI, 1.55–4.70]) of cardiovascular events, in particular coronary events. There was no significant association with all-cause mortality. This study represents the first validation of the clinical significance of the supernormal vascular aging concept, based on prospective data. Its further characterization may help discovering novel protective molecular pathways and providing preventive strategies for successful vascular aging.

Chatterjee S., Budidha K., Kyriacou P.A.

Photoplethysmography (PPG) is a photometric technique used for the measurement of volumetric changes in the blood. The recent interest in new applications of PPG has invigorated more fundamental research regarding the origin of the PPG waveform, which since its discovery in 1937, remains inconclusive. A handful of studies in the recent past have explored various hypotheses for the origin of PPG. These studies relate PPG to mechanical movement, red blood cell orientation or blood volume variations. Objective Recognising the significance and need to corroborate a theory behind PPG formation, the present work rigorously investigates the origin of PPG based on a realistic model of light-tissue interactions. Approach A three-dimensional comprehensive Monte Carlo model of finger-PPG was developed and explored to quantify the optical entities pertinent to PPG (e.g. absorbance, reflectance, and penetration depth) as the functions of multiple wavelengths and source-detector separations. Complementary to the simulations, a pilot in vivo investigation was conducted on eight healthy volunteers. PPG signals were recorded using a custom-made multiwavelength sensor with an adjustable source-detector separation. Main results Simulated results illustrate the distribution of photon-tissue interactions in the reflectance PPG geometry. The depth-selective analysis quantifies the contributions of the dermal and subdermal tissue layers in the PPG wave formation. A strong negative correlation (r = -0.96) is found between the ratios of the simulated absorbances and measured PPG amplitudes. Significance This work quantified for the first time the contributions of different tissue layers and sublayers in the formation of the PPG signal.

Panwar M., Gautam A., Biswas D., Acharyya A.

This paper presents a deep learning model 'PP-Net' which is the first of its kind, having the capability to estimate the physiological parameters: Diastolic blood pressure (DBP), Systolic blood pressure (SBP), and Heart rate (HR) simultaneously from the same network using a single channel PPG signal. The proposed model is designed by exploiting the deep learning framework of Long-term Recurrent Convolutional Network (LRCN), exhibiting inherent ability of feature extraction, thereby, eliminating the cost effective steps of feature selection and extraction, making less-complex for deployment on resource constrained platforms such as mobile platforms. The performance demonstration of the PP-Net is done on a larger and publically available MIMIC-II database. We achieved an average NMAE of 0.09 (DBP) and 0.04 (SBP) mmHg for BP, and 0.046 bpm for HR estimation on total population of 1557 critically ill subjects. The accurate estimation of HR and BP on a larger population compared to the existing methods, demonstrated the effectiveness of our proposed deep learning framework. The accurate evaluation on a huge population with CVD complications, validates the robustness of the proposed framework in pervasive healthcare monitoring especially cardiac and stroke rehabilitation monitoring.

Avram R., Olgin J.E., Kuhar P., Hughes J.W., Marcus G.M., Pletcher M.J., Aschbacher K., Tison G.H.

The global burden of diabetes is rapidly increasing, from 451 million people in 2019 to 693 million by 20451. The insidious onset of type 2 diabetes delays diagnosis and increases morbidity2. Given the multifactorial vascular effects of diabetes, we hypothesized that smartphone-based photoplethysmography could provide a widely accessible digital biomarker for diabetes. Here we developed a deep neural network (DNN) to detect prevalent diabetes using smartphone-based photoplethysmography from an initial cohort of 53,870 individuals (the ‘primary cohort’), which we then validated in a separate cohort of 7,806 individuals (the ‘contemporary cohort’) and a cohort of 181 prospectively enrolled individuals from three clinics (the ‘clinic cohort’). The DNN achieved an area under the curve for prevalent diabetes of 0.766 in the primary cohort (95% confidence interval: 0.750–0.782; sensitivity 75%, specificity 65%) and 0.740 in the contemporary cohort (95% confidence interval: 0.723–0.758; sensitivity 81%, specificity 54%). When the output of the DNN, called the DNN score, was included in a regression analysis alongside age, gender, race/ethnicity and body mass index, the area under the curve was 0.830 and the DNN score remained independently predictive of diabetes. The performance of the DNN in the clinic cohort was similar to that in other validation datasets. There was a significant and positive association between the continuous DNN score and hemoglobin A1c (P ≤ 0.001) among those with hemoglobin A1c data. These findings demonstrate that smartphone-based photoplethysmography provides a readily attainable, non-invasive digital biomarker of prevalent diabetes. A deep neural network applied to smartphone-based vascular imaging can detect diabetes, opening new possibilities for non-invasive diagnosis.

Chen Y., Yang X., Wang D., Liu X., Han X., Zhang J.

Ali S.A., Niaz M.S., Rehman M., Mehmood A., Rahman M.M., Riaz K., Abbasi Q.H.

Hellqvist H., Rietz H., Grote L., Hedner J., Sommermeyer D., Kahan T., Spaak J.

Abstract Wearable technology, such as photoplethysmography (PPG), enables easily accessible individual health data with the potential for improved risk assessment. We hypothesized that the overnight stiffness index (OSI), derived from nocturnal finger PPG, could be used to assess cardiovascular risk and vascular ageing. Subjects with confirmed or suspected hypertension (n = 79, 56 males) underwent simultaneous ambulatory blood pressure monitoring (ABPM) and overnight sleep polygraphy with a continuous PPG registration. Overnight PPG-based pulse propagation time was used to calculate OSI. Associations between OSI and markers of cardiovascular risk, blood pressure, and indices of arterial stiffness, as indicators of vascular ageing, were assessed. Subjects were stratified into low and high OSI (according to median, 10.9 m/s). SCORE2/SCORE2-OP and Framingham risk scores were calculated. The high OSI group had higher SCORE2/SCORE2-OP (9.5 [5.5;12.5] vs 5.0 [4.0;6.5]), and OSI correlated with SCORE2/SCORE2-OP and Framingham risk score (r s = 0.40 and r s = 0.41; both P < 0.01). Indices of arterial stiffness were increased in the high OSI group including ABPM awake and asleep pulse pressures (59 ± 14 vs 50 ± 9 mmHg, P < 0.01, and 54 ± 14 vs 45 ± 7 mmHg, P < 0.001), and ambulatory arterial stiffness index (0.47 ± 0.12 vs 0.37 ± 0.11, P < 0.001), respectively. OSI correlated with 24-h and asleep pulse pressure also after adjusting for confounders. OSI was related to systolic ABPM (awake r = 0.42, asleep r = 0.55; both P < 0.001) and diastolic ABPM (asleep r = 0.36, P < 0.01). OSI, a novel PPG-based measure of nocturnal arterial stiffness, correlates with established cardiovascular risk scores and with blood pressure-derived indices of vascular ageing. This simple method may facilitate cardiovascular risk assessment using readily available medical and wearable consumer devices.

Montero Muñoz N., Tárraga López P.J., López-González Á.A., Paublini H., Martorell Sánchez C., Marínez-Almoyna Rifá E., Ramírez-Manent J.I.

Introduction: The assessment of cardiovascular risk has traditionally relied on validated scales designed to estimate the likelihood of experiencing a cardiovascular event within a specific timeframe. In recent years, novel methodologies have emerged, offering a more objective evaluation of this risk through indicators such as vascular age (VA) and heart age (HA). Objective: This study aimed to investigate the relationship between sociodemographic factors, lifestyle behaviors, and their impact on VA and HA. Materials and Methods: A dual study design, encompassing both cross-sectional and longitudinal retrospective approaches, was conducted among a cohort of employees. The variables assessed included sociodemographic characteristics (age, sex, and socioeconomic status) and health-related habits (smoking, physical activity, adherence to the Mediterranean diet, and alcohol consumption). Results: The findings revealed that all analyzed variables were significantly associated with elevated VA and HA values. Among these, age demonstrated the strongest association, with odds ratios (OR) of 114.91 (95% CI: 100.45–131.43) for high HA and 34.48 (95% CI: 31.41–37.56) for high VA. Conclusions: The profile of individuals most at risk for elevated VA and HA encompasses males of advanced age, characterized by low socioeconomic status, a sedentary lifestyle, poor adherence to the Mediterranean diet, and regular alcohol consumption.

Khalid S., Quraishi I.S., Nawaz M.W., Sajjad H., Yaseen H., Mehmood A., Rahman M.M., Abbasi Q.

Abstract Objective. We study the changes in morphology of the photoplethysmography (PPG) signals–acquired from a select group of South Asian origin–through a low-cost PPG sensor, and correlate it with healthy aging which allows us to reliably estimate the vascular age and chronological age of a healthy person as well as the age group he/she belongs to. Approach. Raw infrared PPG data is collected from the finger-tip of 173 apparently healthy subjects, aged 3–61 years, via a non-invasive low-cost MAX30102 PPG sensor. In addition, the following metadata is recorded for each subject: age, gender, height, weight, family history of cardiac disease, smoking history, vitals (heart rate and SpO2). The raw PPG data is conditioned and 62 features are then extracted based upon the first four PPG derivatives. Then, correlation-based feature-ranking is performed which retains 26 most important features. Finally, the feature set is fed to three machine learning classifiers, i.e. logistic regression, random forest, eXtreme Gradient Boosting (XGBoost), and two shallow neural networks: a feedforward neural network and a convolutional neural network. Main results. For the age group classification problem, the ensemble method XGboost stands out with an accuracy of 99% for both binary classification (3–20 years vs. 20+ years) and three-class classification (3–18 years, 18–23 years, 23+ years). For the vascular/chronological age prediction problem, the ensemble random forest method stands out with a mean absolute error of 6.97 years. Significance. The results demonstrate that PPG is indeed a promising (i.e. low-cost, non-invasive) biomarker to study the healthy aging phenomenon.

Scandelli A., Lombardi C., Castiglioni P., Merati G., Villa F., Faini A.

Amado-Rey A., GonçalvesSeabra A., Stieglitz T.

Cheung M.Y., Sabharwal A., Cot´e G.L., Veeraraghavan A.

Zanelli S., Agnoletti D., Alastruey J., Allen J., Bianchini E., Bikia V., Boutouyrie P., Bruno R.M., Climie R., Djamaleddine D., Gkaliagkousi E., Giudici A., Gopcevic K., Grillo A., Guala A., et. al.

Abstract Vascular ageing (vascular ageing) is the deterioration of arterial structure and function which occurs naturally with age, and which can be accelerated with disease. Measurements of vascular ageing are emerging as markers of cardiovascular risk, with potential applications in disease diagnosis and prognosis, and for guiding treatments. However, vascular ageing is not yet routinely assessed in clinical practice. A key step towards this is the development of technologies to assess vascular ageing. In this Roadmap, experts discuss several aspects of this process, including: measurement technologies; the development pipeline; clinical applications; and future research directions. The Roadmap summarises the state of the art, outlines the major challenges to overcome, and identifies potential future research directions to address these challenges.

Liang T., Yilmaz G., Soon C.

Cardiovascular diseases are a major cause of mortality worldwide. Long-term monitoring of nighttime heart rate (HR) and heart rate variability (HRV) may be useful in identifying latent cardiovascular risk. The Oura Ring has shown excellent correlation only with ECG-derived HR, but not HRV. We thus assessed if stringent data quality filters can improve the accuracy of time-domain and frequency-domain HRV measures. 92 younger (<45 years) and 22 older (≥45 years) participants from two in-lab sleep studies with concurrent overnight Oura and ECG data acquisition were analyzed. For each 5 min segment during time-in-bed, the validity proportion (percentage of interbeat intervals rated as valid) was calculated. We evaluated the accuracy of Oura-derived HR and HRV measures against ECG at different validity proportion thresholds: 80%, 50%, and 30%; and aggregated over different durations: 5 min, 30 min, and Night-level. Strong correlation and agreements were obtained for both age groups across all HR and HRV metrics and window sizes. More stringent validity proportion thresholds and averaging over longer time windows (i.e., 30 min and night) improved accuracy. Higher discrepancies were found for HRV measures, with more than half of older participants exceeding 10% Median Absolute Percentage Error. Accurate HRV measures can be obtained from Oura’s PPG-derived signals with a stringent validity proportion threshold of around 80% for each 5 min segment and aggregating over time windows of at least 30 min.

Jose A.S., Srivastav S., Mehta B.

Objectives: Vascular ageing is increasingly being recognised as a vital marker of cardiovascular morbidity and mortality. Assessment of vascular stiffness is an important parameter in this context. Pulse arrival time (PAT) assessed using photoplethysmography (PPG) and digital electrocardiogram (ECG) signals is a feasible and cost-effective parameter for this assessment. However, there are few, if any, studies that have assessed the test-retest repeatability of this parameter over time. Materials and Methods: We computed PAT using finger PPG and Lead II ECG and measured it sequentially at five instances over a period of 1 month in 21 healthy adults (10 males and 11 females). Mean and diastolic blood pressure (MBP and DBP) and heart rate (HR) were also measured at each visit. A novel parameter, PAT normalised for HR of 75 (PAT-75), was also computed. PAT and PAT-75 were compared for these visits using repeated measures analysis of variance. The intraclass correlation coefficient (ICC) was used to assess the test-retest reliability of this parameter. Results: MBP, DBP, and PAT values did not show any difference between the visits. HR was significantly different between the visits. PAT-75 was significantly lower for the afternoon of day 1 as compared to the forenoon. ICC demonstrated only moderate reliability of PAT (ICC = 0.57), with further reduction observed for PAT-75 (ICC = 0.38). Conclusion: PAT was only moderately repeatable on repeated evaluation over a 1-month period. This finding may have implications for the large-scale applicability of this technology, and therefore, we propose further investigation into the repeatability of this parameter in large cohorts.

Mekonne B.K., Lu W., Hsieh T., Chu J., Yang F.

Cuffless blood pressure (BP) measurements have long been anticipated, and the PPG (Photoplethysmography)-only method is the most promising one since already embedded in many wearable devices. To further meet the clinical accuracy requirements, PPG-only BP predictions with personalized modeling for overcoming personal deviations have been widely studied, but all required tens to hundreds of minutes of personal PPG measurements for training. Moreover, their accurate test periods without calibration practice were not reported. In this work, we collected records of PPG data from our recruited subjects in real-life scenarios instead of relying on the openly available MIMIC dataset obtained from intensive care unit (ICU) patients. Since our objective is commercial application and a substantial reduction in training data, we tailored our model training to closely mimic real-world usage. To achieve this, we developed a training approach that only requires 9-minutes of personal PPG signal recordings and mixed with other PPG data from our recruited 364 subjects. The modeling is conducted with two-channel paired inputs to the convolutional neural network (CNN)-based model, which we called Mixed Deduction Learning (MDL). The test results of 88 samples from 15 subjects, under testing period up to 30-plus days without extra calibration, revealed that MDL meets most of the standards of AAMI, BHS, and IEEE 1708–2014 (for static test only) for BP measurement devices, which indicates MDL’s long-term stability and consistency. Furthermore, we found that the model with two-channel inputs presents a trend of improving performance as the pool of mixed training data increased, while the conventional one-channel input revealed degraded performance. The outperformance of MDL is attributed to many significant features remained in the first CNN layer even when mixing personal 9-minutes data with the other 364 subjects. Consequently, PPG-only with MDL introduces a new avenue for overcoming challenges in training due to personal physiological variations. Given our consideration of real-life usage, this technology can be seamlessly translated to commercial applications.

Nuñez A.P., Ipar E., Armentano R.L., Cymberknop L.J.

Ball J.D., Davies A., Gurung D., Mankoo A., Panerai R., Minhas J.S., Robinson T., Beishon L.

AbstractPrevious studies report contradicting age‐related neurovascular coupling (NVC). Few studies assess postural effects, but less investigate relationships between age and NVC within different postures. Therefore, this study investigated the effect of age on NVC in different postures with varying cognitive stimuli. Beat‐to‐beat blood pressure, heart rate and end‐tidal carbon dioxide were assessed alongside middle and posterior cerebral artery velocities (MCAv and PCAv, respectively) using transcranial Doppler ultrasonography in 78 participants (31 young‐, 23 middle‐ and 24 older‐aged) with visuospatial (VST) and attention tasks (AT) in various postures at two timepoints (T2 and T3). Between‐group significance testing utilized one‐way analysis‐of‐variance (ANOVA) (Tukey post‐hoc). Mixed three‐way/one‐way ANOVAs explored task, posture, and age interactions. Significant effects of posture on NVC were driven by a 3.8% increase from seated to supine. For AT, mean supine %MCAv increase was greatest in younger (5.44%) versus middle (0.12%) and older‐age (0.09%) at T3 (p = 0.005). For VST, mean supine %PCAv increase was greatest at T2 and T3 in middle (10.99%/10.12%) and older‐age (17.36%/17.26%) versus younger (9.44%/8.89%) (p = 0.004/p = 0.002). We identified significant age‐related NVC effects with VST‐induced hyperactivation. This may reflect age‐related compensatory processes in supine. Further work is required, using complex stimuli while standing/walking, examining NVC, aging and falls.

Pal R., Rudas A., Kim S., Chiang J.N., Barney A., Cannesson M.

Detection of the dicrotic notch (DN) within a cardiac cycle is essential for assessment of cardiac output, calculation of pulse wave velocity, estimation of left ventricular ejection time, and supporting feature-based machine learning models for noninvasive blood pressure estimation, and hypotension, or hypertension prediction. In this study, we present a new algorithm based on the iterative envelope mean (IEM) method to detect automatically the DN in arterial blood pressure (ABP) and photoplethysmography (PPG) waveforms. The algorithm was evaluated on both ABP and PPG waveforms from a large perioperative dataset (MLORD dataset) comprising 17,327 patients. The analysis involved a total of 1,171,288 cardiac cycles for ABP waveforms and 3,424,975 cardiac cycles for PPG waveforms. To evaluate the algorithm's performance, the systolic phase duration (SPD) was employed, which represents the duration from the onset of the systolic phase to the DN in the cardiac cycle. Correlation plots and regression analysis were used to compare the algorithm against marked DN detection, while box plots and Bland-Altman plots were used to compare its performance with both marked DN detection and an established DN detection technique (second derivative). The marking of the DN temporal location was carried out by an experienced researcher using the help of the 'find_peaks' function from the scipy Python package, serving as a reference for the evaluation. The marking was visually validated by both an engineer and an anesthesiologist. The robustness of the algorithm was evaluated as the DN was made less visually distinct across signal-to-noise ratios (SNRs) ranging from -30 dB to -5 dB in both ABP and PPG waveforms. The correlation between SPD estimated by the algorithm and that marked by the researcher is strong for both ABP (R2(87,343) =0.99, p= -12 dB for PPG waveforms indicating robust performance in detecting the DN when it is less visibly distinct. Our proposed IEM- based algorithm can detect DN in both ABP and PPG waveforms with low computational cost, even in cases where it is not distinctly defined within a cardiac cycle of the waveform ('DN-less signals'). The algorithm can potentially serve as a valuable, fast, and reliable tool for extracting features from ABP and PPG waveforms. It can be especially beneficial in medical applications where DN-based features, such as SPD, diastolic phase duration, and DN amplitude, play a significant role.