Apparatus for determining a user's cardiovascular risk score

Abstract

Apparatus and methods are provided for determining a cardiovascular risk score from at least cardiovascular data. [Solution] The method includes providing a device adapted to measure a cardiovascular signal of a user, measuring the cardiovascular signal using the device during an observation period having a duration of at least 24 hours, each observation period having a duration of 24 hours having a plurality of observation periods, determining a cardiovascular value for each observation period, compiling the determined cardiovascular values ​​for corresponding measurement periods of each observation period into a set of cardiovascular parameters, creating a 24-hour circadian plot of the set of parameters against the corresponding measurement periods, determining a physiological parameter of the user using the circadian plot, and calculating a cardiovascular risk score for the user using the determined physiological parameter.

Description

[Technical field]

[0001] The present invention relates to an apparatus and method for determining a user's cardiovascular risk score. [Background technology]

[0002] Hypertension remains a leading risk factor for mortality worldwide. Despite its prevalence, blood pressure (BP) management efforts have yet to find success, with the challenge lying in the tool still used to diagnose, measure, and treat hypertension: the sphygmomanometer, invented by Samuel Siegfried Carl von Bach in 1867. In recent years, there has been an explosion of devices that attempt to provide cuff-less (cuffless) estimates of BP, overcoming many of the limitations of cuff-based BP monitors. Unfortunately, the fundamental technological differences between traditional BP cuffs and newer cuffless devices, as well as the reluctance to change well-implemented standards, have led to understandable skepticism and reluctance to adopt cuffless BP monitors in clinical practice.

[0003] The scale of the hypertension problem is difficult to comprehend. Over 1.3 billion people worldwide are estimated to have high blood pressure. Yet, current standards of care for diagnosing, treating, and monitoring hypertension (HTN) result in low rates of control. Continuing with the status quo carries great risks. Hypertension has been the leading cause of cardiovascular disease mortality and morbidity for over 40 years, is estimated to cost the U.S. healthcare system $131 billion annually, and is the leading preventable risk factor for premature death worldwide. Any changes that could improve the treatment of hypertension should be carefully considered, as the potential benefits are significant.

[0004] The past two decades have seen a dramatic digital shift in healthcare, driven primarily by the widespread adoption of electronic health records (HEs), along with a multitude of digital health applications, enabling more robust, real-time, and actionable exchange of data and information between patients and healthcare providers. In fact, twenty years ago (as of the time of this filing) and seven years before the first iPhone was released, the Institute of Medicine (IoM) authored a book highlighting the potential of computer-assisted (now called digital health or mobile health (mHealth)) instruments to automate the transfer of clinical data to clinicians, with the goal of improving clinical care and furthering the understanding of disease.

[0005] Twenty years later, technology is beginning to fulfill the IoM's visionary insight and call to action. Cuffless blood pressure monitoring technology promises to improve the diagnosis, treatment, and monitoring of hypertension, potentially benefiting millions of hypertensive patients. While innovation has long focused on treating the disease after it has developed, technology now offers the opportunity to prevent uncontrolled hypertension and all its attendant risks. The commercialization of cuffless blood pressure monitoring devices will enable the resolution of many of the behavioral and practical challenges in treating a chronic disease that have long remained in the shadows. The treatment of hypertension has been constrained by the limitations of office-based and home cuff blood pressure monitoring. With large-scale adoption, the benefits of cuffless blood pressure monitoring devices can be realized, which will ultimately improve hypertension management worldwide.

[0006] In 1948, one of the most important studies of cardiovascular risk was initiated in Framingham, Massachusetts. The groundbreaking nature of the insights gained over the next 30 years is unquestionable. As part of the study protocol, blood pressure measurements in the clinic were determined by auscultation of Korotkoff sounds, the only technique available at the time. Subjects sat reclining in a chair and only the left arm was measured. To this day, all major guidelines recommend that blood pressure measurements be performed as in the original Framingham study, although there has been a notable shift from manual mercury sphygmomanometers to automated oscillometric devices.

[0007] After decades of using cuff-measured targets in clinical trials and practice guidelines, this method of blood pressure measurement has been established by consensus in the medical community as the standard for estimating blood pressure. Moreover, this method supports the hypothesis that individuals have a physiologically stable and predictable blood pressure. Expert consensus documents such as the American Heart Association (AHA) guidelines also state that "conventional clinical measurements, when measured correctly, are surrogate markers of a patient's true blood pressure, considered as long-term averages, and are considered the most important factor in determining adverse blood pressure outcomes."

[0008] However, as our understanding of hypertension improves, it is becoming clear that blood pressure changes and adapts continuously around the clock in response to changes in lifestyle, daily activities, medications, physical and mental stressors, and body position. Expert guidelines suggest that office measurements should be confirmed with subsequent outside-office measurements over the course of weeks to months, meaning the true nature of an individual's blood pressure pattern in their daily life cannot be fully estimated by relaxing for five minutes with both feet on the floor in a temperature-controlled office, in a quiet environment free of exercise, conversation, caffeine, and noise.

[0009] Ambulatory automated blood pressure monitoring (ABPM) shows some of this variability, but usually only over a 24-hour period. The recent development of cuffless, continuous automated blood pressure monitoring devices has, for the first time, made it possible to capture an individual's blood pressure more accurately over time.

[0010] The importance of out-of-office BP monitoring has been recognized in all major hypertension guidelines that validate office BP measurements. ABPM has been considered the gold standard for out-of-office BP monitoring. However, ABPM remains underutilized for a variety of reasons. Although ABPM may be the current recommended tool for out-of-office BP monitoring, it is rarely used in the United States. Among Medicare beneficiaries, where the prevalence of hypertension is estimated at 50%, only about 0.1% of them use ABPM annually. In China, only 1.6% of surveyed primary care providers reported using ABPM to diagnose hypertension. A simpler, cheaper, and more widely available solution for BP monitoring would be of great benefit to both healthcare professionals and patients.

[0011] Home blood pressure monitoring (HBPM) is recommended in all major hypertension guidelines as an essential adjunct to the diagnosis, monitoring, and management of hypertension. However, in practice, it is difficult for patients to monitor their blood pressure at home and transmit meaningful data. In addition, patients need to be trained to measure blood pressure according to the same standard procedures as blood pressure measurement in a clinic. Despite the ready availability of relatively inexpensive home blood pressure cuffs, the rate of active home blood pressure monitoring in real life is surprisingly low. Half of hypertensive patients report never measuring their blood pressure at home, 10% measure their blood pressure less than once a month, and only 24% of hypertensive patients measure their blood pressure at least once a week. Data show that although HBPM is routinely recommended by expert panels and consensus guidelines, in practice most hypertensive patients do not perform HBPM, and none perform it twice a day for at least 7 days as recommended.

[0012] There are many possible explanations for the significant discrepancy between recommendations and actual practice. A 2017 study explored barriers to primary care providers recommending HBPM to their patients. More than two-thirds of respondents cited one or more of the following reasons as a barrier to obtaining HBPM data: Patients unable to complete the HBPM due to reasons such as low health literacy, time constraints, invasiveness of the test, need for routine, and the need to bring the HBPM into the office. Inaccurate test results can result from patients not adhering to HBPM protocols (e.g., using incorrect cuff size, taking blood pressure measurements at the wrong time, not recording measurements, or "cherry-picking" normal blood pressure values ​​to show the doctor). Inaccurate results due to factors such as the patient's body type. The test results or cuff inflation may increase patient anxiety and therefore may be less accurate.

[0013] Blood pressure (BP) is known to vary over time and to follow a circadian rhythm. Systolic and diastolic values ​​are usually higher during the day and decrease at night and during sleep. This nocturnal drop in BP is commonly referred to as the "nocturnal dipping". However, the exact profile of this nocturnal drop varies widely between individuals and may also change slowly over time. Some parameters of this profile, such as the amplitude of the nocturnal dipping and the morning ascent slope (also called the "morning surge"), are clinically relevant as they correlate with various cardiovascular risk factors. [Prior art documents] [Patent documents]

[0014] [Patent Document 1] European Patent Application Publication No. 3226758 Summary of the Invention

[0015] The present disclosure relates to a method for determining a cardiovascular risk score of a user, the method comprising: To provide a device adapted to measure at least a signal of the cardiovascular system of a user. Measuring at least cardiovascular signals during an observation period having a duration of at least 24 hours, the observation period being further divided into at least one observation interval, the at least one observation interval having a duration of 24 hours and comprising a plurality of measurement periods. Determining cardiovascular values ​​for each of a number of measurement periods within at least one observation period. For each of at least one observation period, the cardiovascular values ​​determined during a given measurement period are collected into a set of cardiovascular parameters. For each of at least one observation period, a 24-hour period of cardiovascular parameters should be plotted (graphed or graphed) against the corresponding measurement period. From the circadian plot, several physiological parameters of the user are calculated. Calculating the user's cardiovascular risk score from the determined (calculated) physiological parameters.

[0016] The present disclosure further relates to an apparatus for determining a cardiovascular risk score of a user.

[0017] The present disclosure further relates to a computer program comprising instructions which, when executed by a computer, cause the computer to carry out the steps of the above method.

[0018] Exemplary embodiments are disclosed in the description of this document and illustrated by the following drawings. [Brief description of the drawings]

[0019] [Figure 1] FIG. 1 ((a) through (d)) illustrates an apparatus and method for calculating a user's cardiovascular risk score. [Diagram 2] 2 to 6 show examples of fitting models that may be employed, the example in FIG. 2 being a trapezoidal model. [Diagram 3] Figures 2 to 6 show examples of fitting models that may be employed, the example in Figure 3 being a rectangular model. [Figure 4] 2 to 6 show examples of fitting models that may be employed, the example in FIG. 4 being a Gaussian model. [Diagram 5] 2 to 6 show examples of fitting models that may be employed, the example in FIG. 5 being a skewed Gaussian model. [Figure 6] 2 to 6 show examples of fitting models that may be employed, the example in FIG. 6 being a skewed squashed Gaussian model. [Figure 7] Figure 7 compares the mean daytime blood pressure values ​​from the device with home blood pressure monitoring (HBPM) blood pressure measurements taken on the same day when initializing the device. [Figure 8]Figures 8(a) and (b) show the low intrasubject reproducibility of ambulatory blood pressure monitoring (ABPM) studies. Because an individual's circadian blood pressure variability changes over time, the selection of arbitrary measurement days in ABPM studies may generate phenotypes that misrepresent a patient's underlying blood pressure phenotype. [Figure 9] FIG. 9 shows a comparison of ABPM and device estimates of nocturnal dipping in blood pressure. [Figure 10a] FIG. 10a shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10b] FIG. 10b shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10c] FIG. 10c shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10d] FIG. 10d shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10e] FIG. 10e shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10f] FIG. 10f shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10g] FIG. 10g shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10h] Figure 10h shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10i] FIG. 10i shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 10j] FIG. 10j shows a new generation of dynamic BP control metrics that can be generated from data captured from the device. [Figure 11] FIG. 11 illustrates an apparatus for determining a user's cardiovascular risk score, according to one embodiment. [Figure 12] FIG. 12 is a cross-sectional view of an apparatus according to one embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0020] In the present disclosure, for a specific user, based on the blood pressure measurement history over a certain period of time, We created a model of the circadian rhythm of blood pressure, deriving from the model a set of relevant parameters that can be used to assess the user's cardiovascular risk score; Further described are methods and devices for communicating this cardiovascular risk score to a user.

[0021] In particular (see Figs. 1A-1B), a cardiovascular signal of a user is measured during an observation period Tm (Fig. 1B) using a device 10 adapted to measure such a cardiovascular signal. The observation period Tm has a duration of at least 24 hours and is divided into a number of measurement periods (not shown). For each measurement period, a cardiovascular parameter of the user is determined.

[0022] Each cardiovascular parameter determined for a corresponding measurement period is collected into a cardiovascular parameter group.

[0023] Graphical representations of the circadian variations of cardiovascular parameters against the corresponding measurement periods are then constructed.

[0024] For a given user, 24-hour, 7-day (24-hour, 7-day) systolic and diastolic blood pressure values ​​are collected together for a continuous period (e.g., 2 weeks). Regarding the time element of each measurement, the exact date of each measurement is ignored, and only the time information (hours and minutes) is stored. This process combines all blood pressure measurements (systolic or diastolic, respectively) for the period under study into a single 24-hour period centered around midnight. Blood pressure measurements that fall outside the hourly measurement period (e.g., from 7:00 AM to 7:59:00 AM), defined as the median ± 2 × IQR (interquartile range) for that period, are discarded (outlier rejection) before further analysis.

[0025] A constrained piecewise linear model is fitted to the remaining data by least-squares optimization, in which the geometrical features (profiles) of the BP are constrained to be constant during the nighttime slope and to another constant value during the day, allowing two linear transitions (ramps) between these two states.

[0026] Once constructed, this model generates a set of six physiological parameters that uniquely describe an individual's BP profile (see Figure 1b). These parameters are: Diurnal BP value (Y 0 ), Absolute nighttime amplitude (ampl), Time of start of the ramp (X0), Duration of lamp before night (dl), Duration of the slope plateau (nl), Lamp duration after night (al) It is.

[0027] The physiological parameters may also be combined with additional clinically relevant parameters, such as (non-exhaustively) Nighttime BP value, Relative night-time fluctuation range, Duration of nighttime fluctuations, Time in Target Range (TTR), or Slope of morning surge It is.

[0028] The calculated physiological parameters may further determine the user's blood pressure phenotype, including classifications such as true normal blood pressure, white coat hypertension, masked hypertension, sustained hypertension, hypotension, nocturnal drop, and nocturnal rise.

[0029] Physiological parameters, clinically relevant parameters and / or blood pressure phenotypes may also be incorporated into the additional value of cardiovascular risk.

[0030] In the embodiment shown in Figures 1a to 1d, the method for calculating a user's cardiovascular risk score comprises the following steps. Providing a device 10 adapted to measure at least a signal of the cardiovascular system of a user (FIG. 1a); measuring said at least cardiovascular signals using the device 10 during an observation period Tm having a duration of at least 24 hours (FIG. 1b), measuring a cardiovascular signal, the observation period Tm being divided into at least one observation interval Ts, where the at least one observation interval Ts lasts twenty-four (24) hours and comprises a plurality of measurement periods; determining a cardiovascular value for each measurement period of a plurality of measurement periods within said at least one observation period Ts; collecting the cardiovascular values ​​determined for a given measurement period of each of said at least one observation period Ts into a group of cardiovascular parameters; constructing, for each of at least one observation interval Ts, a 24-hour circadian plot of the parameters against the corresponding measurement period; determining a plurality of physiological parameters of the user using the circadian plot (FIG. 1c); calculating a cardiovascular risk score for the user using the determined physiological parameters (FIG. 1d); Equipped with.

[0031] The measurement period corresponds to the period during which the cardiovascular signals are being measured by the device 10 .

[0032] In Figures 1a to 1d, FIG. 1a shows a device 10 embodied as a cuffless wrist blood pressure optical sensor; FIG. 1b shows a cardiovascular value (in this case a blood pressure value) calculated from a cardiovascular signal (in this case a PPG signal, not shown) provided to the device 10; Figure 1c shows the collection of cardiovascular values ​​for each observation interval Ts over the entire observation period Tm (see Figure 2) into a group of cardiovascular parameters, the construction of a 24-hour circadian plot, fitting a model to the circadian plot, and the determination of multiple physiological parameters from which the user's cardiovascular risk score is calculated (Figure 1d).

[0033] In one aspect, the observation period Tm has a duration of 48 hours, 7 days, 1 month, or 1 year. Note that if the observation period Tm has a duration of 24 hours, its duration is equal to the duration of the observation interval Ts.

[0034] In another aspect, the measurement period may have a duration of at least 10 seconds or 30 seconds, or 1 minute, 5 minutes, 1 hour, 2 hours, 4 hours, or 6 hours.

[0035] In the example of Fig. 1b, the observation period Tm is more than 24 hours. For example, the observation period Tm corresponds to 2 weeks (14 days). In that case, the observation period Tm comprises one or more observation intervals Ts (14 observation intervals lasting 24 hours for an observation period Tm of 2 weeks).

[0036] Each 24-hour observation period Ts comprises a number of measurement periods, for example, 24 measurement periods lasting one hour, and thus a number of cardiovascular values ​​are calculated for each measurement period corresponding to an hour of the day (24 cardiovascular values ​​are calculated for each hour of the day).

[0037] Thus, multiple cardiovascular values ​​are collected into cardiovascular parameter sets for each time period of the day corresponding to each measurement period (see FIG. 1c).

[0038] In one aspect, cardiovascular signals may be measured separately on weekdays and non-weekdays.

[0039] Cardiovascular signals may include any that measure the quantity and quality of sleep, values ​​derived from electrocardiogram signals, values ​​derived from photoplethysmography signals, values ​​derived from bioimpedance signals, values ​​derived from ultrasound signals, or values ​​derived from arterial pulsation signals such as pulse pressure, or any that measure the degree of physical activity.

[0040] The cardiovascular value is calculable from the cardiovascular signal measured during the measurement period. Calculating the cardiovascular value may comprise identifying pulses of the cardiovascular signal. Calculating the cardiovascular value may further comprise determining at least one characteristic of the identified pulses and calculating the cardiovascular value based on the at least one characteristic. Calculating the cardiovascular value may further comprise determining at least one characteristic of the identified pulses and calculating the cardiovascular value based on the at least one characteristic. For example, a method for calculating the cardiovascular value is presented in US Pat. No. 6,399,633. In other embodiments, the cardiovascular value may be calculated by applying a non-parametric or machine learning algorithm without identifying multiple consecutive pulses of the cardiovascular signal.

[0041] The step of determining the cardiovascular value may comprise calculating a cardiovascular value from each cardiovascular signal. The cardiovascular value may comprise any one of a systolic blood pressure value, a diastolic blood pressure value, a mean blood pressure value, a heart rate value, or a blood glucose value.

[0042] Any of the following cardiovascular parameters (pulse pressure, central pulse wave velocity, peripheral pulse wave velocity, arterial stiffness, aortic pulse transit time, augmentation index, cardiac output, cardiac output parameters, pulse pressure parameters, systemic vascular resistance, venous pressure, systemic hemodynamic parameters, pulmonary hemodynamic parameters, cerebral hemodynamic parameters, heart rate, heart rate variability, beat-to-beat interval, arrhythmia detection, ejection time, SpO 2(oxygen concentration) saturation), SpHb (total hemoglobin concentration), SpMet (methemoglobin concentration), SpCO (carboxyhemoglobin concentration), respiratory rate, tidal volume, general cardiovascular and health indicators).

[0043] The step of collecting cardiovascular values ​​may comprise grouping the cardiovascular values ​​determined by the device 10 during a given measurement period into cardiovascular parameters (see Figures 1b and 1c).

[0044] The step of constructing a 24-hour circadian plot may comprise plotting (charting) the group of cardiovascular parameters calculated for each measurement period as a function of time over a 24-hour period, for example from midnight to the next midnight (see FIG. 1c).

[0045] In one aspect, the cardiovascular parameters may have a duration of 10 seconds, 30 seconds, 1 minute, 5 minutes, 10 minutes, or 30 minutes.

[0046] The step of using the circadian plot may comprise calculating a plurality of physiological parameters of the user from the circadian plot.

[0047] In one aspect, the step of using the circadian plot may comprise calculating, for each cardiovascular parameter group, a cardiovascular representative value of the cardiovascular values ​​collected in that cardiovascular parameter group.

[0048] In another aspect, using the circadian plot comprises calculating a cardiovascular representative, which may comprise classifying the cardiovascular parameters as inliers or outliers and calculating the cardiovascular representative using only the inlier cardiovascular parameters.

[0049] In one aspect, calculating the cardiovascular representative may comprise classifying the cardiovascular parameters as inliers or outliers. The cardiovascular parameters may be weighted according to their probability of being an inlier or outlier.

[0050] In one aspect, determining a plurality of physiological parameters of the user comprises fitting a model to the circadian plot, the physiological parameters corresponding to parameters of the model.

[0051] The model may be one of a linear model, a non-linear model, a constrained model, an unconstrained model, fitted by least squares optimization.

[0052] In one aspect, the constraint model may include time constraints on the duration of a physiological variable (e.g., the nighttime blood pressure dip plateau may not be longer than 12 hours, or the duration of the pre-night slope may not be longer than the duration of the dip plateau), or may include amplitude constraints on a physiological variable (e.g., daytime blood pressure values ​​may not be higher than 200 mmHg, or the morning systolic blood pressure spike may not be greater than 50 mmHg / hour).

[0053] As shown in Figure 2, the physiological parameters were the daytime blood pressure values ​​(Y 0 ), the absolute dip amplitude during the night (ampl), and the time component of the dip onset (X 0 ), duration of the pre-night ramp (dl), duration of the dip plateau (nl), or duration of the post-night ramp (al).

[0054] Figures 2 to 6 show examples of model fitting, namely: Six parameters Y o , X o ,A trapezoidal model including ampl, dl, nl, and al (Fig. 2); Four parameters Y o , X o ,A rectangular model including ampl and nl (Fig. 3), Four parameters Y o , X o ,Gaussian model (Fig. 4) including sd,ampl. Five parameters Y o , X o , s.d. o , s.d.1 ,A skewed Gaussian model with ampl (Fig. 5), Six parameters Y o , X o , sdo, sd 1 , ampl, and nl, which is a skewed flattened Gaussian model (Figure 6).

[0055] The physiological parameters may include differences between physiological parameters calculated on weekdays and weekends, or, more generally, differences between physiological parameters calculated on weekdays and non-weekdays.

[0056] The physiological parameters may further include any temporal dynamics of the physiological parameters, which may include the day-to-day variation of the parameter, the variability of the parameter over days, the trend of the parameter over days, or the number of days that the parameter is above or below a threshold.

[0057] In one aspect, the model further uses non-physiological parameters, including parameters such as: Your geographic location, including your altitude; weather forecasts and observations, heat waves, cold waves, travel information, pollution information, and public health information, including the outbreak of infectious diseases in your location; Season, allergy information, sunshine pattern, social information, education level, family situation (marital status, number and age of children), Financial information, political interests and views, professional circumstances, including degree of responsibility, type of contract, regularity of working hours; Information from the calendar (including business days / non-business days and holidays, workload, and work schedules), dietary patterns, exercise and activity patterns, sedentary levels, Leisure information, including caffeine, alcohol, and drug use; Medicines (not limited to antihypertensives), the health condition of the user and their close relatives, Type and size of residence, whether you own or rent, whether you have pets, Surveys on social media use, religious practices, mood and general state It is.

[0058] The method may further comprise combining at least two physiological parameters to obtain one or more related physiological parameters. The combining may include adding, multiplying, or dividing the at least two physiological parameters. Such combining may further include calculating a correlation coefficient or a synchronization coefficient of the two or more physiological parameters.

[0059] Relevant physiological parameters include any of the following: Daytime blood pressure value, nighttime blood pressure value, relative nighttime dip amplitude, duration of nighttime dip, time in target range (TTR), Blood pressure fluctuation pattern, morning surge slope, nighttime systolic blood pressure (SBP), nighttime diastolic blood pressure (DBP), nighttime heart rate (HR), SBP fall, DBP fall, HR fall, SBP morning surge, DBP morning surge, Morning surge in HR, time of fall in SBP, time of fall in DBP, time of fall in HR, synchrony of SBP or DBP or HR, Response to medication type, antihypertensive medication compliance, indicators of patient initiative, or response to lifestyle modification instructions It is.

[0060] In one aspect, the relevant physiological parameters may include the user's blood pressure phenotype.

[0061] In one aspect, the relevant physiological parameters include any of the following: True normal blood pressure, white coat hypertension, masked hypertension, sustained hypertension, hypotension, nocturnal hypotension, nocturnal hypertension, or A phenotype that predicts response to a particular drug or treatment (such as renal denervation).

[0062] The step of calculating a cardiovascular risk score of the user using the determined physiological parameters may comprise calculating a cardiovascular risk score of the user from the determined physiological parameters.

[0063] In one aspect, calculating the user's cardiovascular risk score comprises using user data.

[0064] User Data may include any of the following: Age, weight, height, sex, ethnicity, lipid levels, diabetes status, smoking, CT calcium (Agaston score), Family history, genetic markers for disease risk, actigraphy information (information from actigraphs (smartwatch-type small, highly sensitive acceleration sensors and loggers)), exercise information, dietary information, Stress levels, general feeling, hormone data, menstrual cycle information, Medication information, weekday or weekend information, seasonal information, sleep quality information, sleeping patterns, Any of the parameters used to calculate cardiovascular risk scores in clinical guidelines such as the American College of Cardiology / American Heart Association (ACC / AHA) guidelines, the European Society of Cardiology (ESC) guidelines, or the Multi-Ethnic Atherosclerosis Study (MESA) database; The user data may be any non-physiological parameter.

[0065] User data may be provided manually by users or may be automatically integrated from external systems.

[0066] The cardiovascular risk score may be the 10 year risk of cardiovascular disease, the 10 year risk of heart disease, the 10 year risk of stroke, or any other clinically relevant cardiovascular risk score.

[0067] In one aspect, the cardiovascular parameters may be at least any of a blood pressure value, a heart rate value, a cardiac output value, a blood glucose value, a measure of physical activity, or a measure of sleep amount and quality.

[0068] Also disclosed herein is a device for determining a user's cardiovascular risk score.

[0069] The device 10 (see FIG. 11) comprises a measurement module 20 adapted to measure signals of the user's cardiovascular system during an observation period Tm lasting at least 24 (twenty-four) hours, the observation period Tm being divided into a number of observation intervals Ts lasting 24 hours each having a number of measurement periods.

[0070] The device 10 further comprises a processing unit 30 adapted to determine cardiovascular values ​​from the measured cardiovascular signals for each measurement period, compile the determined cardiovascular values ​​for the corresponding measurement periods of each observation interval Ts into a group of cardiovascular parameters, construct a circadian plot in which the group of 24-hour parameters is plotted against the corresponding measurement periods, determine a plurality of physiological parameters of the user using the circadian plot, and calculate a cardiovascular risk score for the user using the determined physiological parameters.

[0071] The device further comprises an interface 40 for displaying and / or transmitting the calculated cardiovascular risk score. The interface 40 is operatively connected to the processing unit 30.

[0072] The processing device 30 may be configured to perform the steps of determining cardiovascular values, collecting cardiovascular values, generating a 24-hour circadian plot, calculating a number of physiological parameters of the user, and calculating a cardiovascular risk score of the user.

[0073] The device 10 may be operatively connected to a wired or wireless communication link, the latter of which may include WiFi or Bluetooth or cellular support (mobile data communication). The device 10 may further be operatively connected to a memory.

[0074] The interface 40 may comprise an application on a smartphone, tablet, computer, smartwatch or any mobile device.

[0075] The interface 40 may be adapted to generate an external signal to the user to provide guidance on how to optimize the user's calculated abnormality risk, for example by suggesting lifestyle, medication or treatment changes.

[0076] The interface 40 may be adapted for inputting non-physiological or user data, either manually or automatically.

[0077] The interface may be located near the user or far away from the user.

[0078] In one aspect, the measurement module 20 may be adapted to automatically measure cardiovascular signals without user interaction.

[0079] The measurement module 20 includes the following: Galvanic Skin Response (GSR) sensor arrays, Bio-Impedance (BioZ) sensor arrays, Electrocardiogram sensors (ECG), sensors based on Radio Frequency (RF) detection, radar sensors, Mechanical sensors, pressure sensors, invasive sensors, intra-arterial sensors, minimally invasive sensors, Subcutaneous sensor, tonometer, strain sensor, pulse wave sensor, microphone, Ultrasonic sensors, Capacitive sensors, Electromagnetic sensors, Raman sensors, Any sensor capable of measuring pulsatile signals from the capillary bed of the skin or any other part of the arterial tree. The present invention may include any one of the arterial pulsation sensors. Thus, the cardiovascular signals measured by the device 10 via the measurement module 20 may correspond to signals measured (by any of the sensors mentioned above).

[0080] The device 10 may comprise a wearable device. A possible configuration when the device 10 is a wearable device is shown in cross section in Fig. 12. The device 10 may include a wristband 15 with a measurement module 20. The measurement module 20 may comprise at least one pulse sensing unit 21. For example, the measurement module 20 comprises four pulse sensing units 21 arranged along the inner periphery of the wristband 15 so as to contact the skin of the user's wrist when the device 10 is worn. The pulse sensing units 21 may be arranged on the wristband 15 in other ways.

[0081] In one embodiment, the pulse sensing unit 21 may comprise a photoplethysmography (PPG) sensor array that measures arterial pulsation, arterial diameter, blood flow, and / or blood constituents. In this case, the cardiovascular signal is a photoplethysmography (PPG) signal. In this embodiment, the pulse sensing unit 21 may be mounted on the wristband 15 such that the optical sensor array 21 spans an artery such as the ulnar artery 111 (near the ulna 113) or the radial artery 112 (near the radius 114 or any arterial vascular bed 117 in the skin of the wrist), or otherwise.

[0082] In one embodiment, the apparatus 10 further comprises a trigger module 50 (see FIG. 11) configured to start or stop a measurement period by the measurement module 20 .

[0083] The trigger module 50 can control the measurement module 20 according to trigger parameters. The trigger parameters can be user specific. An example of a trigger parameter can be a trigger signal, such as a motion signal representative of a user's movement. Such a motion signal can be measured, for example, using a motion sensor 60 disposed on the device 10. The motion sensor 60 can include any of an inertial measurement unit (IMU), an accelerometer, a gyroscope, a geomagnetometer, or a combination of these devices.

[0084] In one aspect, the cardiovascular parameter is a blood pressure (BP) value.

[0085] Blood pressure can be measured using cuff-based BP technology and / or optical sensors. Blood pressure is measured at the wrist.

[0086] The blood pressure value comprises at least one of systolic BP, diastolic BP or mean BP.

[0087] The device 10 disclosed herein allows for the delivery of blood pressure values ​​continuously, for example throughout the day and night.

[0088] The device 10 disclosed herein represents a cuffless blood pressure monitoring device. Cuffless blood pressure devices solve many practical and behavioral problems and overcome the barriers to recommended routine monitoring of blood pressure, and have the potential to capture significantly more blood pressure data than traditional methods. The ability to collect continuous blood pressure measurements at home, outside the workplace, during daily activities, while asleep, etc. over weeks, months, or years can provide patients and healthcare providers with a more complete blood pressure assessment than intermittent testing in controlled positions and environments.

[0089] Traditionally, monitoring blood pressure outside the workplace and continuously throughout the day and night has been limited by available monitoring techniques that require the cuff to be inflated each time a measurement is taken. Devices that do not require cuff inflation could overcome most of the limitations inherent in traditional cuff-based blood pressure monitors.

[0090] The device 10 allows for the blood pressure of an individual to be determined non-invasively, without causing arterial occlusion.

[0091] The device 10 may be placed on locations on the body such as the wrist, fingertip, chest, ear, forehead, or combinations thereof.

[0092] The device 10 provides an indirect blood pressure estimate that relies on analysis of the arterial pulse at one or more body locations with a sensor that does not apply pressure at that location. The device 10 does not provide a direct pressure measurement, but rather supplies a quantity that a computer program calculates from analysis of the waveform of the pressure pulse that, following an initialization phase, is usually correlated to a blood pressure value. From optical sensors (assessing cutaneous arteriolar pulsation via reflectance or transmission photoplethysmography sensors) A camera sensor (to assess skin fibrillation using video-reflected photoplethysmography); Biopotential sensors (to assess different electromagnetic signatures of cardiac activity or to assess arterial pulsatility from impedance plethysmography signals at different body sites); Radar sensors (which evaluate the arterial pulsations in different body parts from radar reflections), Multiple sensor technologies (sensor unit 21) ranging from acoustic pressure sensors (which assess the pulsatility of superficial arteries by sensing skin displacement) are currently used to capture the pressure pulse waveform. Depending on the number of body sites on the patient's body where the pressure pulse is captured, the analysis of the waveform is based either on a Pulse Wave Velocity algorithm (usually when at least two body sites are involved) or on a Pulse Wave Analysis algorithm (usually when only one body site is involved). As the evaluation of such pulse waveforms does not involve pressure measurements, most cuffless blood pressure monitors require an initialization procedure that uses an oscillometric device to provide information in "mmHg".

[0093] A major attraction of the device 10 is the potential to provide many more blood pressure data points. The ability to collect blood pressure values ​​continuously over days to years at home, outside the workplace, during daily activities, overnight and while asleep provides patients and healthcare providers with a much more representative assessment of blood pressure than occasional cuff estimates. The blood pressure values ​​measured at a single point in time, either in a clinic or via home blood pressure monitoring, are only a small part of the entire dynamic data set of blood pressure. Without these (the entire dynamic data set) data, doctors and patients are essentially clueless about the true nature of blood pressure.

[0094] As an illustration of the above limitations, actual systolic blood pressure (SBP) data from a male subject was recorded over a two month period using device 10. The data shows significant discrepancies between simulated office measurements, occasional home blood pressure monitoring, and blood pressure monitoring using a portable automated blood pressure monitor. Office estimates suggest absolute systolic blood pressure values ​​that are significantly higher than average and do not capture blood pressure data over time, which is most clearly and commonly seen in white coat and masked hypertension syndromes (up to 40% of subjects). Home blood pressure measurements, when taken routinely, may correlate with overall average values ​​but do not capture day-to-day or circadian variations. Finally, portable automated blood pressure monitors only reveal information for a limited period of time, such as 24 or 48 hours. All three conventional methods of blood pressure measurement capture only a small portion of the dynamic blood pressure, which is very well demonstrated by device 10.

[0095] Compared to other existing technologies, the device 10 can provide significantly more blood pressure measurements, better indicate blood pressure variability, and measure nocturnal blood pressure, all of which are clinically meaningful. However, clinicians may wonder how the daytime average blood pressure provided by the device 10 compares to home blood pressure monitoring (HBPM) measurements taken on the same day. To answer this question, we performed an offline analysis of anonymized data from 2928 users of the device 10 (Figures 7a and 7b). The analysis compared daytime blood pressure data (8:00 am to 8:00 pm) measured with the device 10 on the day of the initialization procedure with data measured by upper arm cuff blood pressure measurement (as home blood pressure monitoring HBPM) during the same initialization procedure. The analysis was repeated for both systolic and diastolic blood pressure. In both cases, the difference between the measurements by the two methods was shown to be statistically significant by paired T-tests (both p<0.001). However, the difference (2.25 mmHg systolic and 0.44 mmHg diastolic) was below the resolution and margin of error of the blood pressure measured by the automated blood pressure monitoring device.

[0096] Figures 7a and 7b show the average daytime systolic and diastolic blood pressure values ​​from the device 10 compared to HBPM measurements taken during device initialization on the same day. Figure 7a shows the systolic and diastolic blood pressure values ​​for 2928 users of the device 10, each with one data point. The X-axis is the single measurement taken with the HBPM during device initialization, and the Y-axis is compared to the average daytime blood pressure (8 am to 8 pm) measured with the device 10 on the same day. On the left are the systolic blood pressure values ​​(SBP) and on the right are the diastolic blood pressure values ​​(DBP). The dotted lines are calculated using Huber linear regression. The numerical plots in Figure 7b show the distribution of the differences between the values ​​measured by device 10 and the corresponding HBPM values ​​for systolic and diastolic blood pressure. The mean and standard deviation of each distribution are shown at the bottom of the figure. An asterisk indicates a statistically significant difference between the values ​​measured by HBPM and device 10 (both p<0.001).

[0097] Using real-world data from over 2000 patients using the device 10, the remarkably rich blood pressure data set, the ability to measure nocturnal blood pressure longitudinally, and the automated and passive nature of the device stand out as distinct advantages when compared to traditional monitoring methods (Table 1). Furthermore, the systematic differences between the device 10 daytime mean blood pressure and the HBPM daytime measurements in this population are small and within the acceptable margin of error. In summary, the device 10 has great potential to significantly improve the ability to monitor blood pressure in outpatients. [Table 1] Table 1. Guidance for interpreting daily mean values ​​from Instrument 10 in comparison with those from HBPM.

[0098] Although its use in clinical practice is limited (see previous section), ambulatory blood pressure monitoring (ABPM) remains the recommended method when a more complete analysis of a patient's blood pressure patterns is required for the diagnosis / monitoring of hypertension. ABPM is currently the only recommended method and allows for day and night blood pressure measurements, but it is underused and its reproducibility has been questioned. Circadian blood pressure variations are known to be dynamic, and arbitrarily short observation periods (24 or 48 hours) may provide clinicians with data that are representative of only a small portion of the overall blood pressure.

[0099] To further illustrate the issue of ABPM reproducibility, Figures 8a and 8b show two examples of repeated ABPM recordings from an ongoing clinical trial (bottom panel). A meta-analysis of 35 observational studies has shown that for one-third of patients, the classification as dipper or non-dipper is not reproducible across two consecutive nights of ABPM measurements (e.g., a patient is classified as a dipper on the first night, but in the non-dipper group on the next night) and that the difference in nighttime mean systolic and diastolic blood pressures over two consecutive nights can vary from -19.6 to 21.3 mmHg and -11.3 to 12.3 mmHg, respectively. These data support that ABPM has limited reproducibility in assessing intra-individual dipping status and daytime and nighttime blood pressure values. Figures 8a and 8b show ABPM recordings from selected patients from the ABPM study. FIG. 8a shows very poor within-subject reproducibility, while FIG. 8b shows better reproducibility in measuring nocturnal diastolic blood pressure and daytime and nocturnal mean blood pressure.

[0100] Given the documented poor reproducibility of ABPM, the device 10 may overcome the arbitrariness of cuff-based ABPM by leveraging the device 10's ability to generate large amounts of data over long periods of time. However, two factors must be considered when comparing data from the two methods. First, the device 10 measures blood pressure in a fundamentally different way than that measured by traditional oscillometric ABPM. Second, the frequency and periodicity of ABPM measurements (e.g., every 20 minutes) differs from that of the device 10 (e.g., when the user is stationary for a long enough period of time). The differences capture blood pressure during different daily activities, with ABPM capturing more measurements during physically active periods than device 10. Thus, the differences between the two (technical and activation timing) produce average values ​​of daytime and nighttime blood pressure that may vary depending on the measurement method.

[0101] 9a to 9d provide a first overview of the systematic differences observed between ABPM and the device 10 in estimating a patient's nocturnal dipping state.

[0102] In Figures 9a and 9b the difference in estimated SBP reduction in patients from the NCT04548986 study is shown. Figure 9c shows a systemic factor of 3.4 across the initial patient cohort of the same study. Figure 9d shows a systemic factor of 3.1 across N=4644 users of the device 10 when matched to the phenotypic distribution of a large independent study (N=6359). As expected, the different observational modalities tend to provide similar phenotypic information, but the application of technology-dependent conversion factors is required.

[0103] Figures 9a and 9b show an example of concurrent BP data acquired by an ABPM monitor (Diasys 3 Plus, Novacor, France, Figure 9a) and device 10 (Aktiia BP monitor, Aktiia, Switzerland, Figure 9b) from one patient enrolled in the NCT04548986 trial. Note that the ABPM data was recorded over a 24-hour period, while the data from device 10 was recorded during and for one week after the ABPM recording. The observation period with the device 10 was extended to one week to take into account the day-to-day variations in circadian patterns and to increase the number of data points registered during the day and night due to the low sampling frequency of the device 10. On the same plot, two estimates of blood pressure dip were extracted. To calculate the dip on ABPM recordings, the common approach was used: the difference between daytime and nighttime blood pressure. Daytime and nighttime subperiods were defined based on fixed clock time intervals: daytime from 9am to 9pm and nighttime from 12am to 6am. Data recorded during the transitional periods were excluded to avoid too much variance between individual users. Data points located at a distance higher than the interquartile range from the median of each subperiod were considered outliers, and finally, the dip was calculated as the difference between the medians of the subperiods. A statistical approach was implemented to calculate the dips recorded by the device 10. Taking advantage of the fact that the cuffless circadian plot has a high density of data points, a parametric model was used to fit the circadian rhythms of SBP and DBP (see the "fit" line in the plot series). The night-time dip was then extracted from one of the parameters of the model. The estimated night-time dip in this patient already differs in two ways: the ABPM dip (A) appears to be larger than the cuffless dip (B).

[0104] Figure 9c shows the statistical analysis of the A>B phenomenon in the preliminary patient population of the NCT04548986 study. Data from the 20 initially enrolled patients were processed to estimate the systematic gain difference comparing dipping measured by ABPM with dipping measured by the device10. In this cohort, after bootstrapping the available samples, ABPM dipping was found to be characteristically 3.4 times greater than cuffless dipping, with a 95% confidence interval ranging from 2.3 to 4.4. It is important to note that NCT04548986 is not yet completed and a comprehensive analysis of the collected data will be presented in a further dedicated publication.

[0105] Figure 9d shows a further statistical investigation of the same A>B phenomenon, combining data from an independent ABPM study of N=6359 patients with real-world data from N=4644 users of the device 10. According to Kario et al.'s study, the nocturnal BP phenotype (a phenotype is defined as an observable characteristic of the circadian rhythm of BP) measured by ABPM measurements in the general population is expected to show the following distribution: 16% of individuals were extreme dippers (a drop of more than 20mmHg); 40% of individuals were normal dippers (a drop of 10 to 20 mmHg); 32% of individuals were non-dippers (a drop from 0 to 10 mmHg); 12% of individuals are risers (positive dips). However, when observing the dip patterns recorded by the device 10, the same distribution is not met and the dip distribution is clearly compressed. In this analysis, we estimated the optimal factor required to expand the dip distribution of the device 10 to a representation equivalent to that of ABPM. In this cohort of N=4644 users, after applying bootstrap methods, we found that the dips from ABPM were characterized by a factor of 3.1 greater than the dips from cuffless continuous measurements. The 95% confidence interval ranges from 2.8 to 3.4.

[0106] By integrating real-world data from over 4,000 users of the device 10 with clinical data from controlled clinical trials, it has been demonstrated that the device 10 can estimate a patient's blood pressure phenotype. However, due to technical differences between the cuffless technology of the device 10 and oscillometric ABPM monitors, technology-dependent conversion factors may be required to compare estimates from both methods. Table 2 summarizes guidelines on how blood pressure phenotype data from the device 10 should be interpreted when compared to blood pressure characteristics derived from ABPM.

[0107] Thus, the device of the present invention may further comprise converting the calculated cardiovascular values, calculated cardiovascular parameters, calculated circadian plots or calculated physiological parameters into ABPM-equivalent values ​​and plots according to a mapping function, which may be a pre-calculated affine function as described above (where the ABPM-like dip can be calculated as 3.1 times the dip calculated by the device 10), or any other type of mapping function pre-calculated from data recorded from a large population, or calculated from any of the user data as described in claim 15, or a combination of both. The same mapping approach may be applied to convert a calculated cardiovascular value, a calculated cardiovascular parameter, a calculated circadian plot, or a calculated physiological parameter into an ABPM equivalent value and plot according to a mapping function, and the ABPM equivalent or HBPM equivalent may then be used in place of or in addition to the associated physiological parameter in calculating the user's cardiovascular risk score. [Table 2] Table 2. Guidelines for interpretation of BP phenotype data obtained from the device 10 when compared with BP characteristics obtained from ABPM.

[0108] To illustrate the potential performance of the device 10, FIGS. 10a to 10J show a new set of dynamic BP control indices estimated for a male patient (51 years old) during 5 months of continuous observation with the device 10.

[0109] In particular, Figures 10a through 10J illustrate a new generation of dynamic BP control metrics that can be generated from data captured from device 10. The time series presented were captured from subjects over a five month period. In addition to standard BP indices such as 24-h, daytime and nighttime averages of SBP, DBP and HR (Figs. 10a, 10c, 10e, 10f and 10g), we present new dynamic indices such as time to therapeutic range (TTR, Fig. 10b), SBP variability (Fig. 10d), dynamic circadian model (Fig. 10h), dynamic nighttime SBP decline (Fig. 10i) and dynamic nighttime decline duration and morning surge acceleration (Fig. 10j).

[0110] Figure 10a reports all 4729 SBP measurements taken during the observation period and the evolution of the 24-h mean SBP. SBP measurements within the target range (<120 mmHg) are shown as green dots, and those outside the target range (>120 mmHg) are shown as red dots. Based on these data, Figure 10b shows a new metric for SBP in BP control: time in target range SBP (Y-axis quartiles: 75% to 100% time = green, 50% to 75% time = yellow, 25% to 50% time = orange, 0% to 25% time = red) with the total percentage of time spent in each quartile on the right. Figure 10c further displays the nighttime mean SBP, with SBP>120mmHg colored in red and SBP<120mmHg colored in green. Figure 10d displays the mid-term blood pressure variability (SD in mmHg) during the day (orange), night (black) and mean (green). Figures 10e and 10f show the 24-hour average diastolic blood pressure (green line) and heart rate (green dots) as well as all the individual data points. Figure 10g displays the daytime (orange), nighttime (black), and 24-h average (green) SBP. The shaded time periods a, b, and c in the same figure correspond to Fig. 10h and are also highlighted in panels I and J. Thus, Fig. 10h shows the circadian pattern of SBP during time periods a, b, and c, which show different patterns of nocturnal decline, duration of the nocturnal decline, and morning surge in this patient. Regarding sleep-related blood pressure variables, Figure 10i shows the evolution of the SBP night-time decline in mmHg (dark green) and % (light green). Figure 10j also shows the evolution of the duration of the nocturnal SBP decline (red) and quantifies the morning surge (orange).

[0111] The advent of 24-h ABPM demonstrated that the BP phenotype is more complex than a simple binary variable (hypertensive or not) and allowed the demonstration of diurnal variations in BP, including daytime and nighttime BP components and the nocturnal physiological drop in BP. The predictive value of these different components was compared, and each component showed predictive value independent of absolute BP. Nighttime mean BP (systolic or diastolic) was a stronger predictor of cardiovascular events. The predictive value of nocturnal mean BP has been shown to be superior to 24-h ABPM, daytime mean BP, and HBPM in both hypertensive and general populations. This high predictive value is noteworthy because the reproducibility of the dipping pattern has been shown to be low in small and recent larger studies. Indeed, only a small proportion of hypertensive patients maintain their initial dipping phenotype for more than 4 years. In addition, nocturnal ABPM is less well tolerated than daytime, which may affect sleep quality. Sleep disturbances caused by cuff inflation have also been shown to affect the association of nocturnal ABPM with outcomes.

[0112] The device 10 has the potential to overcome these unfavorable characteristics of nocturnal ABPM. First, nocturnal blood pressure can be measured repeatedly over days to months, allowing a more consistent nocturnal phenotype to be derived. Second, by not inflating the cuff, the impact on sleep quality is expected to be negligible. Nevertheless, these unique characteristics may affect the normal value of nocuffless blood pressure at night, which may need to be redefined.

[0113] To encourage exploration of new phenotype-driven assessments of cardiovascular risk, Table 3 provides a list of existing promising blood pressure phenotypes that can be driven by large-scale deployment of the device 10. [Table 3] Table 3. List of recommendations for BP phenotypes already identifiable by HBPM / ABPM screening and list of expansions of BP phenotypes that will be enhanced with the deployment of the device 10.

[0114] The utility and predictive value of classical and new phenotypes must be demonstrated in longitudinal epidemiological studies before we can be confident of their use in daily practice. Independent predictive values ​​for 24-h mean or nocturnal BP will be necessary if evidence-based personalized treatment is to become a reality. Studying these phenotypes will take time, but they are sure to provide physicians with a panel of new physiological or induced BP responses that may help personalize antihypertensive treatment in the future.

[0115] The present disclosure further relates to a non-transitory computer-readable storage medium comprising a computer program product including instructions for causing at least one processing device to perform a method for determining a cardiovascular risk score of a user.

Claims

1. Providing a device (10) adapted to measure at least a cardiovascular signal of a user; measuring said at least cardiovascular signals during an observation period (Tm) having a duration of at least 24 hours, The observation period (Tm) is divided into at least one observation section (Ts), - measuring said at least one cardiovascular signal, said at least one observation interval (Ts) having a duration of 24 hours and comprising a plurality of measurement periods; determining a cardiovascular value for each of said plurality of measurement periods within said at least one observation period (Ts); - collecting, for each of said at least one observation period (Ts), cardiovascular values ​​determined during a given measurement period into a set of cardiovascular parameters; generating a 24-hour circadian plot of cardiovascular parameters for a corresponding measurement period for each of said at least one observation period (Ts); calculating a plurality of physiological parameters of the user from said circadian plot; calculating a cardiovascular risk score for said user from the determined physiological parameters; A method for providing the above.

2. 2. The method of claim 1, wherein the observation period (Tm) has a duration of 48 hours, 7 days, 1 month or 1 year.

3. 3. The method according to claim 1 or 2, wherein the measurement period has a duration of at least 10 seconds or 30 seconds, i.e. 10 seconds, 30 seconds, 1 minute, 5 minutes, 1 hour, 2 hours, 4 hours or 6 hours.

4. The method further comprising:

4. The method of claim 1, comprising calculating for each group of cardiovascular parameters a cardiovascular representative value of the cardiovascular values ​​collected in said group of cardiovascular parameters.

5. Calculating said representative value of the cardiovascular system classifying said cardiovascular parameter as an inlier or an outlier; calculating said representative value of the cardiovascular system using only said inlier cardiovascular system parameters; The method of claim 4 comprising:

6. Calculating said representative value of the cardiovascular system classifying the cardiovascular parameter as an inlier or an outlier; The method of claim 5 , wherein the cardiovascular parameters are weighted according to their probability of being an inlier or an outlier.

7. Determining a plurality of physiological parameters of the user 7. The method of claim 1, comprising fitting a model to the circadian plot, wherein a plurality of said physiological parameters correspond to a plurality of parameters of the model.

8. The method of claim 7 , wherein the model is one of a linear model, a nonlinear model, a constrained model, and an unconstrained model that is fitted by least squares optimization.

9. The physiological parameters are daytime blood pressure values ​​(Y ０ ), the absolute dip amplitude during the night (ampl), the time element of the dip onset (X ０ 9. The method of claim 1, further comprising any one of the following: duration of the pre-night ramp (dl), duration of the dip plateau (nl), or duration of the post-night ramp (al).

10. The method of claim 7 , wherein the model further uses a number of non-physiological parameters.

11. The method of claim 1 , further comprising combining at least two of the physiological parameters to obtain one or more related physiological parameters.

12. The one or more relevant physiological parameters include: Daytime blood pressure value, nighttime blood pressure value, relative nighttime dip amplitude, duration of nighttime dip, time in target range (TTR), Blood pressure fluctuation pattern, morning surge slope, nighttime systolic blood pressure (SBP), nighttime diastolic blood pressure (DBP), nighttime heart rate (HR), SBP decreased, DBP decreased, HR decreased, SBP morning spike, DBP morning spike, Morning surge in HR, time of decline in SBP, time of decline in DBP, time of decline in HR, synchrony of SBP or DBP or HR, Response to medication type, antihypertensive medication compliance, indicators of patient initiative, or response to lifestyle modification instructions The method of claim 11, comprising any one of:

13. The method of claim 11 or 12, wherein the relevant physiological parameters include a blood pressure phenotype of the user.

14. The relevant physiological parameters include: Phenotypes predicting true normotension, white coat hypertension, masked hypertension, sustained hypertension, hypotension, nocturnal dip, nocturnal hypertension, or response to specific drugs or therapies The method of claim 13, comprising any one of:

15. 15. The method of claim 1, wherein calculating the cardiovascular risk score for a user comprises using user data.

16. The user data includes: Age, weight, height, sex, ethnicity, lipid levels, diabetes status, smoking, CT calcium (Agaston score), Family history, genetic markers for disease risk, actigraphy information, exercise information, dietary information, Stress levels, general feeling, hormone data, menstrual cycle information, Medication information, weekday or weekend information, seasonal information, sleep quality information, sleeping patterns, Any of the multiple parameters used to calculate cardiovascular risk scores in clinical guidelines such as the American College of Cardiology / American Heart Association (ACC / AHA) guidelines, the European Society of Cardiology (ESC) guidelines, or the Multi-Ethnic Atherosclerosis Study (MESA) database; 16. The method of claim 15, comprising any one of:

17. 17. The method of any one of claims 1 to 16, wherein the cardiovascular risk score is a 10-year risk of cardiovascular disease, a 10-year risk of heart disease, a 10-year risk of stroke, or any other clinically relevant cardiovascular risk score.

18. 18. The method according to any one of claims 1 to 17, wherein the cardiovascular parameters are at least one of the following: blood pressure values, heart rate values, cardiac output values, blood glucose values, measures of physical activity, measures of sleep amount and quality, electrocardiogram signals, photoplethysmography signals, bioimpedance signals, ultrasound signals.

19. a measurement module (20) adapted to measure signals of the user's cardiovascular system during an observation period (Tm) lasting at least 24 hours, the observation period (Tm) being divided into a number of observation intervals (Ts), each observation interval (Ts) lasting 24 hours with a number of measurement periods (Tm); A processing device (30) adapted to determine a cardiovascular value for each measurement period (Tm), the processing device compiles the determined cardiovascular values ​​for each corresponding measurement period (Tm) of each observation interval (Ts) into a set of cardiovascular parameters; generating a 24-hour circadian plot of said cardiovascular parameters versus said corresponding measurement period (Tm); using said circadian plot to determine a plurality of physiological parameters of a user; a processing unit (30) for using the determined physiological parameters in calculating a cardiovascular risk score for the user; an interface (40) for displaying and / or transmitting the calculated cardiovascular risk score; 1. An apparatus for determining a cardiovascular risk score of a user, comprising:

20. The apparatus of claim 19, wherein the interface (40) comprises a smartphone, a tablet, a computer, a smartwatch, or a mobile device.

21. 21. A device according to claim 19 or 20, wherein the device is connectable to a wired communication line or to a wireless communication line including WiFi or Bluetooth or cellular support.

22. A non-transitory computer readable storage medium comprising a computer program product comprising instructions for causing at least one processing unit to perform the method of any one of claims 1 to 18.

Images