Device for determining one or more parameters indicating a condition of a heart of a living being

Abstract

Various aspects relate to a device configured to: receive reception data representing a time series of radar reception signals having a sampling rate equal to or greater than 100 Hz; for each range bin of a monitored area, determine, for each point in time of the time series a respective amplitude value and a respective phase value; for each object, determine respective one or more range bins are associated with the object; determine, whether at least one object of the one or more objects is a living being; and in the case that it is determined that at least one object is a living being, determine, using a temporal change of the respective phase value of at least one range bin of the respective one or more range bins associated with the at least one object, one or more parameters indicating a condition of a heart of the living being.

Description

DEVICE FOR DETERMINING ONE OR MORE PARAMETERS INDICATING A CONDITION OF A HEART OF A LIVING BEINGTechnical Field

[0001] Various aspects relate to a (sensor) device for determining one or more parameters indicating a condition of a heart of a living being.Background

[0002] The continuous assessment of health and well-being, particularly at home, can be an effective means of detecting early signs of the onset of ailments so that early medical intervention may be offered, thereby reducing the need for hospitalization and in many cases avoiding it altogether.

[0003] In view of the increasing nursing shortage due to a lack of staff in the healthcare system, the resulting shortage of beds in the inpatient sector and the ever-increasing life expectancy of people and thus the ageing of the population, the need for continuous monitoring in both the inpatient and home sectors is becoming increasingly important.

[0004] Continuous monitoring of various health parameters that indicate the state of health may be used to detect diseases at an early stage or to make a reasonable risk assessment for a person who, for example, develops chronic heart failure or the like.

[0005] Continuous monitoring of older people in their familiar home environment can also help them to remain independent for longer, as relatives or the emergency services may be informed immediately in an emergency to help the person.

[0006] Therefore, it may be desirable to provide a (indoor) monitoring system that allows to detect various parameters indicating the state of health of a person.

[0007] There are various parameters indicating a condition of a heart of a living being. Some of these parameters are only hardly (e.g., using invasive methods) detectable, such as a heart rate variability, a vascular elasticity, etc.Brief Description of the Disclosure

[0008] Various aspects relate to a (sensor) device that is capable to contactless (viz. not invasively) acquire various parameters which indicate a condition of a heart of a living being, such as a heart rate variability, a (aortic) blood pressure, a vascular elasticity, a ratio between a (aortic) systolic period and a (aortic) diastolic period of the heart, a (aortic) systolic blood pressure, a (aortic) diastolic blood pressure, etc.

[0009] This is achieved by using a radar detection approach described herein in combination with a comparatively high sampling rate.Brief Description of the Drawings

[0010] In the drawings, like reference characters generally refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead generally being placed upon illustrating the principles of the invention. In the following description, various aspects of the invention are described with reference to the following drawings, in which:FIG.1 shows a system for detecting one or more parameters indicating a condition of a heart of a living being according to various aspects;FIG.2 shows a diagram illustrating, at a position associated with a skin surface of a human, a phase value over time when using a sampling rate of 20 Hz, the diagram indicating a movement of the skin resulting from breathing and heart rate;FIG.3 shows the Wiggers diagram for illustrating various aspects of cardiac physiology;FIG.4 shows a comparison of the phase value over time curve profile between a sampling rate of 100 Hz and a sampling rate of 1000 Hz according to various aspects;FIG.5 shows a measured heart cycle using a sampling rate of 1000 Hz; andFIG.6 shows the measured skin movement on different positions on a human according to various aspects.Description

[0011] The following detailed description refers to the accompanying drawings that show, by way of illustration, specific details, and aspects in which the invention may be practiced. These aspects are described in sufficient detail to enable those skilled in the art to practice the invention. Other aspects may be utilized and structural, logical, and electrical changes may be made without departing from the scope of the invention. The various aspects are not necessarily mutually exclusive, as some aspects can be combined with one or more other aspects to form new aspects.

[0012] FIG.l shows a system 100 for detecting one or more parameters indicating a condition of a heart of a living being (e.g., a human, a person, or an animal, etc.) according to various aspects. The system 100 may include a device 106. The system 100 may include one or more radio-frequency (RF) transceivers 102. Each of the one or more RF transceivers 102 may be configured to transmit radar transmit signals in direction of the one or more objects(e.g., into a monitored area) and to receive radar reception signals being a version of the radar transmit signals after reflection at the one or more objects (in the monitored area).

[0013] According to various aspects, the device 106 may include the one or more RF transceivers.

[0014] For this the RF transceiver may include a phase-controlled antenna array. The phase-controlled antenna array may include a plurality of (e.g., at least three) transmit antennas and one or more reception antennas. According to various aspects, the RF transceiver may be configured to transmit the RF signals with a frequency of at least 60 GHz (e.g., in a range from about 60 GHz to about 64 GHz; or a greater frequency such as with a frequency of about 120 GHz). According to various aspects, the phase-controlled antenna array may be configured such that the radar beam can be steered into a desired direction. This allows to steer the radar beam to specific portions (e.g., body parts) of the living being, thereby allowing to extract signals with a greater signal-to-noise (SNR) ratio.

[0015] The RF transceiver may use an in-phase and quadrature method to determine the in- phase component and the quadrature component for determining amplitude and phase values, respectively.

[0016] To increase a signal-to-noise ratio (and, thus, provide a better, more exact, and more precise detection of the amplitude and phase values) the radar signals may be repeatedly transmitted and received before the reception data 104 are evaluated.

[0017] The one or more RF transceiver 102 may be configured to continuously transmit radar transmit signals (over time) and receive their reflected versions as radar reception signals. The one or more RF transceivers 102 may be configured to provide these reception data 104 representing a time series of the radar reception signals to the device 106. The device 106 may include one or more processors configured to process the reception data 104. Thus, when referring to a processing carried out by the device 106, it may be understood that this processing may be carried out by the one or more processors.

[0018] According to various aspects, the one or more RF transceiver 102 may be configured to employ a binary phase modulation. This allows to use the plurality of transmit antennas and the one or more reception antennas in parallel (by removing the overlap during processing). Simultaneous transmission improves the signal-to-noise ratio and, thus, the accuracy of the amplitude values and phase values.

[0019] The device 106 may be configured to divide the monitored area (e.g., having a lateral area of 5 meters times 5 meters or less, e.g., 3 meters times 3 meters or less) into a plurality of three-dimensional (3D) range bins. For this, the device 106 may determine distance data using the reception data 104 (e.g., from the received radar reception signals). For example, the monitored area may be divided radially from the device 106 into the range bins. The extension of a range bin may be in a range from about 1 centimeter to about 5 centimeters, preferably about 3 cm. However, the size of each distance section may depend on the bandwidth of the one or more RF transceivers 102. For example, a minimum achievable size of each distance section depends on the bandwidth of the one or more RF transceivers 102.

[0020] According to various aspects, a distance between the device 106 and the living being monitored may be equal to or less than 2 meters (e.g., equal to or less than 1 meter, e.g., equal to or less than 0.8 meters). It has been found that the highest accuracy of the determined one or more parameters indicating the condition of the heart of a living being is achieved when using a distance in a range from about 0.8 meters to about 1.0 meters. A maximum distance that still allows to determine the one or more parameters indicating the condition of the heart of a living being as detailed herein may depend on the number of antennas and a beamwidth.

[0021] The device 106 (e.g., the one or more processors thereof) may be configured to, for each range bin of the plurality of 3D range bins, allocate, for each point in time of the time series, a segment of the radar reception signal to the range bin, and determine a respective amplitude value and a respective phase value using the segment of the radar reception signal.Illustratively, the device 106 may determine a respective amplitude value and a respective phasevalue for each range bin (may also be referred to as range segment or distance segment). This may be done by analyzing a frequency curve of the radar reception signals. The amplitude value may be a measure of a reflection strength of the radar reception signal (viz. of the signals reflected by the one or more objects in the respective range bin). The phase value may be a measure of a distance of an object within a range segment. However, the phase value may not represent an absolute value for a distance, as several distances can correspond to the same phase value due to the small wavelength of the radar reception signals in contrast to the larger distance section. Therefore, only changes in distance can be measured / detected. If an object moves within a range bin, the changes may be detected with an accuracy in the micrometer range, for example with an accuracy of about 30pm or less.

[0022] Illustratively, dividing the monitored area into the plurality of 3D range bins in combination with tracking a change of the phase value within each range bin over time (in 108) allows to determine movements in a micrometer range such as a movement of a skin surface of a living being when breathing and / or due to the heartbeat.

[0023] Optionally, the radar reception signals may be preprocessed to exclude static objects from the radar reception signals since they are irrelevant when analyzing a parameter that indicates condition of a heart of a living being. This allows to simplify and speed up the processing. Therefore, according to various aspects, the radar reception signals may be preprocessed such that objects with different movement intensities can be distinguished. For this, the reception data 104 may be processed using various calculation rules. The resulting data may be saved and then further processed. This allows the results to be combined later. The preprocessing of the reception data 104 may be done, for example, by removing the static objects using a first calculation rule. Static objects are objects that do not exhibit any movement. This makes it easier to distinguish living beings from other objects. The resulting data simplify and speed up a recognition of the living being that is monitored. A second calculation rule may be used to determine the respective (e.g., skin) movements in each rangebin and for each of the one or more objects. This enables mapping of the monitored area and is helpful, for example, for detecting a presence of a living being (e.g., in a bed and / or in a room). A third calculation rule may be used to recombine the resulting data after the first calculation rule over several processing cycles. As a result, fast large movements, e.g. walking or falling of living beings, if present, are deducted and fine slow movements, such as breathing and speaking, are amplified. This is helpful for recognizing creatures when they are behaving very quietly, e.g. while sleeping or after a fall. For example, this information may be used to identify whether there is a living being in the monitored area who has been lying in bed for a long period of time. The pressure ulcer monitoring described herein can then be carried out for this person, for example, by monitoring their position. It is understood that the processing may be carried out for each range bin of the plurality of range bins separately.

[0024] According to various aspects, the respective phase value of each range bin of the plurality of range bins at each time step of the time series may be determined using an extended differentiation and cross-multiplication algorithm. The extended differentiation and crossmultiplication algorithm (Extended DACM) may be used instead of an arctan function for determination of the phase value, which allows a faster determination of the phase values.

[0025] In 110, for each object of the one or more objects, respective one or more range bins of the plurality of -dimensional range bins are determined which are associated with the object. In various aspects, using a point cloud preprocessing as detailed in the following, a point cloud may be determined and one or more points of the point cloud may be allocated to each object of the one or more objects.

[0026] According to various aspects, the device 106 may be configured to subdivide each of the one or more range bins associated with a corresponding object into a plurality of subbins (illustratively employing a zoom-in). This allows to increase the resolution of the object.

[0027] According to various aspects, a respective reflection strength of the radar reception signals may be determined for each range bin of the plurality of range bins at each point in timeof the time series. According to various aspects, each range bin of the plurality of range bins may be additionally divided into angular ranges for the azimuth angle and elevation angle using a spherical coordinate system. Using a spectrum-based beamforming method, the reflection strength may be determined for each of these spatial areas defined by distance and solid angle. Valid reflection values may be determined from the reflection strength for each 3D range bin in the monitored area and a point cloud may be generated therefrom. A valid reflection value may be a signal strength of the radar reception signals that originates from a target object. These can vary greatly, as they depend on the distance and surface material of the object, for example. To avoid this problem, a threshold value for a valid reflection value may be recalculated for each spatial area, which is also referred to as a cell, from its local environment. For this, each cell may be compared with neighboring cells. To do this, a sum of the surrounding cells may be determined for a specific area. Directly adjacent cells may be ignored. The sum may be calculated with a fixed factor to the threshold value. The threshold value may be compared with the value of the cell to be examined. If the value of the cell is greater than the threshold value, this is a valid reflection value and the spatial area is transferred to the point cloud as a point. Empirically, the threshold value may be set to an optimum so that all target objects are detected with a minimum number of false detections. The evaluation of the cells may be carried out in a combination of ID and 2D processing, which is fast and accurate enough, or directly as 3D processing.

[0028] For example, the individual points in the point cloud that have similar values in terms of position, speed, and signal -to-noise ratio may be associated with a same object of the one or more objects. Points can either be assigned to an object known from the previous processing cycle or points can be combined to form a new object. Points that cannot be assigned may be discarded.

[0029] It may be checked which points of the point cloud can be assigned to an existing object. This may include two steps: Firstly, the determination of a threshold value for thedistance and secondly, the determination of a number of points for the affiliation. When determining the distance threshold value, the predicted state vector (location, speed, acceleration) of the object, the previous distribution of points of this object in space and the measurement uncertainty may be taken into account. If a point in the point cloud has a smaller distance than the calculated limit value, it may belong to this object. This may be the case for several objects. The next step may be to determine the number of points for the affiliation to ensure a clear assignment. To do this, the Mahalanobis distance between a point and each possible object may be calculated with increased factorization of the speed. Since the objects to be distinguished are already spatially close to each other, the amplification of the speed during the calculation provides a better assignment. The point is assigned to the object with the highest number of points.

[0030] If points are combined to form a new object, the values of any point may be assumed to be the object center and object speed. All remaining points may be checked one after the other to see whether they comply with the threshold values for distance and speed difference. If the thresholds are met, the new point is added to the object and the object center and object speed are recalculated in order to start the comparison again with the next point. After all existing points have been tested, the created object is checked for validity. According to various aspects, a valid new object may be saved if the threshold values for the number of points, the combined signal-to-noise ratio and the minimum speed are exceeded.

[0031] In general, the respective position of each object in the monitoring area may be tracked so that, for example, an X / Y position in the monitored area can be detected and monitored to detect the presence of the object.

[0032] Furthermore, in addition to the X / Y position in the monitoring area, both spatial and temporal monitoring of a Z position of the object may be detected and monitored. The center of gravity of the point cloud of all points of an object may be the Z-position. However, the minimum / maximum Z-position (lowest / highest point) may also be included in the evaluation.For example, the Z-position varies considerably when walking, as different parts of the body reflect differently in each processing cycle. If the person / object is lying down (e.g. in a bed or on the floor), it does not matter where reflections occur, the center of gravity will always be at a low height.

[0033] As detailed herein, the time dependence of the phase values (viz. the temporal change of the respective phase value) in each range bin may represent object movements in the micrometer range. The skin of the living being may reflect the radar transmit signals and a movement of the skin caused by breathing and / or heartbeat may result in a change of the phase values over time. According to various aspects, the change of the phase values may be directly proportional to a change in distance caused by the skin movement. FIG.2 shows a diagram illustrating, at an example position at a skin surface of a human, a phase value over time when using a sampling rate of 20 Hz.

[0034] The device 106 may be configured to determine, in 112, whether at least one object of the one or more objects is a living being. For this, the device 106 may, in general, use the time series of radar reception signals. The device 106 may be configured to determine that an object of the one or more objects is a living being using the information regarding larger movements of the object within the monitored area and / or may use the phase values of the one or more range bins associated with each object to determine based on the fine-movements (indicated by the phase values) whether the object is a living being. Additionally or alternatively, the device 106 may include a thermal sensor (e.g., a thermopile array) configured to detect a heat signature within the monitored area. The heat signature of each object may be used to determine whether the object is a living being. The heat signature of each object may also be used as a plausibility check of the radar data-based determination whether an object is a living being.

[0035] As shown in FIG.2, there may be an overlap of phase changes induced by breathing and by the heart rate. According to various aspects, the device 106 may be configured to filterthe phase values using phase changes induced by breathing and phase changes induced by the heart rate to determine the blood pressure curve. Various aspects of the processing are detailed in the following.

[0036] After extracting the phase values (e.g., of the whole body of the living being), the breathing rate may be determined using the extracted phase values (e.g., employing one or more filters, one or more fast Fourier transformations (FFTs), and / or a peak analysis). Further, the heart rate may be detected using the extracted phase values (e.g., employing one or more filters, one or more fast Fourier transformations (FFTs), one or more Wavelet transformations (DWTs), and / or a phase correction). When evaluating the spectrum for the heart rate, interference from harmonic frequencies of breathing and random body movements may be taken into account. These also cause high amplitudes in the frequency range of the heart rate. It has been found that the presence of high amplitude values at the harmonic frequencies is a reliable distinguishing feature. A harmonic frequency may be considered to be present if a limit value for the signal-to-noise ratio is exceeded in a frequency range by an integer multiple of the fundamental frequency of an amplitude value. Interference may also have high amplitude values in the heart rate range (0.8 Hz - 2.0 Hz), but no high values in the harmonic range (1.6 Hz - 6.0 Hz). Only if no harmonic frequencies are found, the frequency value may not be a multiple of the respiratory rate. When determining the heart rate, the calculated amplitude spectrum may therefore be cleaned of these interfering harmonic frequencies of respiration.

[0037] According to various aspects, only frequency values that fulfill the following rules may be taken for further data analysis:- If a 1st and 2nd harmonic of the selected frequency value is present, the frequency value may be stored four times,- If the 1st or 2nd harmonic of the selected frequency value is present, the frequency value may be stored three times,- If the amplitude value is the highest in the amplitude spectrum and the associated frequency value is not a harmonic of the breathing frequency, the frequency value may be stored twice,- If a difference between a frequency value for one of the three highest amplitude values in the amplitude spectrum and a result for the heart rate from a previous processing cycle is less than a limit value, this frequency value is accepted.The rules may be checked in the aforementioned order until a rule applies.

[0038] According to various aspects, a cluster analysis may be applied to the stored frequency values. This allows to exclude outliers from the averaging and to suppress interference in individual distance sections. The mean value of the largest cluster may be used. According to various aspects, a median value thereof may be determined over a multiple processing cycles. This median value may be the heart rate.

[0039] According to various aspects, a correlation method may be used to further check the heart rate. For this, the extracted phase values are filtered with one or more bandpass filter. The filters may have a fixed bandwidth (e.g. 2 Hz) and filter the phase values in the harmonic range (2 - 5 Hz). On this new data time series several techniques can be used to enhance the harmonic frequency parts of the heart signal in the temporal domain. These techniques may include standardization, squaring, (partial) scaling, moving sum calculations, derivative filters and peak detection. Afterwards an auto-correlation may be calculated. If the spectrum fulfills certain aspects (number, amplitude ratio and frequency difference of peaks) the first peak may be taken of the spectrum as a heart rate. This heart rate value may be compared with heart rate result from the spectrum analyzing method. If the difference of these values is smaller than a threshold value the result from the spectrum analyzing method may be taken into the output buffer.

[0040] The detected breathing and heart rate may be used to filter the phase signals for aorta blood pressure curve selection. Changes in the phase values resulting from breathing motion, random body movements and noise may be filtered out. An Algorithm (e.g. a Kalman filter)may be used to track the changes in phase values that are caused by breathing of the object. Filtering can be done by subtracting the filtered values from the phase values. Furthermore, Random Body Movements can be filtered by using a derivative filter. The detected heart rate may also used to select / detect phase signals with the aorta blood pressure curve.

[0041] For detection of the blood pressure curve, the filtered phase signals for blood pressure curve selection may be divided into time periods corresponding to the determined heart rate. These times series may be analyzed by a peak detection method. The algorithm may analyze the signal and may compare the signal to predefined identification markers (e.g. number of peaks, amplitude ratios, gradients) that correspond to the curve shape of an aortic blood pressure curve. A confidence value representing the match between markers and signal may be calculated and compared to a threshold value. In another version of the algorithm, a set of predefined curves may be used to do a correlation calculation. In this calculation a correlation coefficient (e.g. pearson correlation) may be calculated between the standardized filtered phase signals and the predefined curves that may be scaled for the corresponding heart rate. The correlation coefficient may be compared to a threshold value.

[0042] Once the blood pressure curve is determined, the one or more parameters 112 may be determined as detailed herein. As described herein, the sampling rate may define which parameters 112 can be determined from the blood pressure curve (depending on their resolution). For this, specific features of the blood pressure curve may be determined, such as the start of the steep rise of the blood pressure curve, the dicrotic notch, an average value, etc. For example, the HRV may be determined using a time interval between the start of the steep rise of the blood pressure curve of two consecutive cycles. For example, the average blood pressure (may also be referred to as mean blood pressure) may be determined. For example, the ratio between the (aortic) systolic period and the (aortic) diastolic period of the heart may be determined using the start of the steep rise of the blood pressure curve and the time of the dicrotic notch (e.g., in both directions).

[0043] The device 106 (e.g., the one or more processors thereof) may be configured to determine, in the case that it is determined that at least one object is a living being, using a temporal change the respective phase value of at least one range bin of the respective one or more range bins associated with the at least one object, one or more parameters 112 indicating a condition of a heart of the living being.

[0044] The device 106 may be configured to provide the one or more parameters 112 and / or information regarding a health condition determined based on the one or more parameters 112 to another (remote) device. Since the reception data 104 do not include any privacy information about the living being (since its reflection signals only) and since the device 106 only outputs the one or more parameters 112 and / or the information regarding the health condition, data privacy is ensured.

[0045] According to various aspects, a respective confidence level may be defined for each parameter of the one or more parameters 112 and may be used to check whether the phase values of the parameter are plausible or whether they are false positive detected. For this, the determined value of each parameter may be saved at each time step and the median of these values may then be determined over a predefined period of time. The number of phase values having a difference to the median equal to or less than a predefined threshold value may be counted. A ratio of this number to the total number of values may define the confidence level. If there are many different values, the confidence is low, otherwise it is high. This has the advantage that values from periods in which the device 106 is probably positioned incorrectly are not included in the averaging.

[0046] FIG.3 shows the Wiggers diagram for illustrating various aspects of cardiac physiology. For example, the Wiggers diagram shows various parameters, such as a temporal change of an aortic pressure, an atrial pressure, a ventricular pressure, a ventricular volume, etc.

[0047] It has been found that various of these parameters can be determined using the radio detection method described herein in combination with a respective threshold sampling rate.

[0048] The respective threshold sampling rate may indicate a threshold sampling rate that is at least required (viz. using a sampling rate equal to or greater than the threshold sampling rate) in order to determine the corresponding parameter.

[0049] It has been found that, for example, a threshold sampling rate of at least 100 Hz is required to determine the heart rate variability (HRV) as a parameter of the one or more parameters 112. The HRV may also be referred to as cycle length variability and / or R-R variability. The HRV is a variation in the time periods between heartbeats. The resolution of the temporal phase value curve resulting from the threshold sampling rate (of equal to or greater than 100 Hz) allows to detect a rising edge associated with a rising edge of the blood pressure curve.

[0050] However, other parameters may require a threshold sampling rate greater than 100 Hz. For example, to determine an average (aortic) blood pressure and / or a vascular elasticity, it has been found that a threshold sampling rate of at least 250 Hz is required. The vascular elasticity may be indirect proportional to the average blood pressure. In particular, it has been found that although a sampling greater than the threshold sampling rate of at least 100 Hz allows to determine the HRV, a sampling rate of equal to or greater than the threshold sampling rate of at least 250 Hz is required in order to detect the blood pressure curve with the peak of systolic phase and diastolic phase. Illustratively, the sampling rate equal to or greater than the threshold sampling rate of at least 250 Hz allows to achieve a resolution of the blood pressure curve that allows to determine an average value and / or to use an estimation formula based on the amplitude values of systolic and diastolic peak.

[0051] It has been found that even a ratio between a (aortic) systolic period of the heart and a (aortic) diastolic period of the heart can be detected when using a threshold sampling rate of at least 500 Hz. To detect the ratio between the (aortic) systolic period and the (aortic) diastolic period, the dicrotic notch (see FIG.3) has to be detected. Using a sampling rate less than the threshold sampling rate of about 500 Hz does not allow to detect the dicrotic notch. This isillustratively shown in FIG.4 which shows a first temporal phase value curve 402 which is detected using a sampling rate of 100 Hz and a second temporal phase value curve 404 which is detected using a sampling rate of 1000 Hz. As illustratively shown, using the sampling rate of 100 Hz (curve 402) does not allow to detect the dicrotic notch 406. The ratio between the (aortic) systolic period and the (aortic) diastolic period may represent a cardiac performance of the heart. An exemplary measurement curve of a heart cycle is shown in FIG.5 (given by a measured skin movement over time) measured using a sampling rate of 1000.

[0052] It has been found that even the (aortic) systolic blood pressure and the (aortic) diastolic blood pressure when using a sampling rate of at least 1000 Hz in combination with detecting a respective temporal phase value curve at two different positions of the living being. Hence, a first temporal phase value curve may be detected at a first (skin) position (e.g., at at least one extremity, such as a foot and / or hand) of the living being and a second temporal phase value curve may be detected at a second (skin) position (e.g., in a thoracic region) (different from the first position; e.g., a positional difference between chest and extremities) of the living being. A blood pressure curve velocity may be determined using a temporal shift between the first temporal phase value curve and the second temporal phase value curve and using a positional difference between the first position and the second position. The blood pressure curve velocity may then be used to determine the blood pressure (e.g., using the Moens- Korteweg equation). It has been found that the sampling rate of at least 1000 Hz allows to determine a difference between a start of the steep rise of first temporal phase value curve (representing a first blood pressure curve) and the second temporal phase value curve (representing a second blood pressure curve) with an accuracy in the millisecond (ms) range. It has been found that each blood pressure curve should be correct to 5 mmHg, resulting in a pulse wave velocity accuracy of about 0.1 m / s. With this, the time difference between the steep rise of the respective blood pressure curve should be correct to 4 ms. With this, the time for the steep rise should be correct to 1ms.

[0053] As detailed herein, various different sampling rate threshold values are required to be able to determine certain parameters. This results from a specific resolution of the blood pressure curve required determine a respective parameter and an increasing sampling rate reduces the SNR and therefore improves the resolution of the blood pressure curve.

[0054] FIG.6 shows the measured skin movement on different positions on a human according to various aspects. As illustratively shown with reference also to the Wiggers diagram shown in FIG.3, measurement curve 602 allows to detect the aortic pressure, measurement curve 604 may allow to detect the ventricular pressure, measurement curve 606 allows to detect the jugular venous pulse.

[0055] It has been found that the aortic pressure can, for example, be detected using the skin movement in the chest region, the head or a leg of a human.

[0056] The (left) ventricular pressure may, for example, be detected using the skin movement in the chest region (e.g., in a region close to the heart).

[0057] It has been found that the atrial pressure and / or jugular venous pulse can, for example, be detected using the skin movement next to the external jugular vein and / or in an upper chest region.

[0058] According to various aspects, the device 106 may be configured to determine a plurality of parameters 112 using various positions on the living being. This may allow to correlate the determined parameters to one another, thereby increasing the accuracy of the determined parameters.

[0059] Besides allowing to detect the one or more parameters 112 contactless, the device 106 also allows to detect the one or more parameters 112 when distanced by several meters (e.g., up to 4 meters) from the living being. Further, the device 106 allow to continuously detect the one or more parameters 112.

[0060] According to various aspects, the device 106 (e.g., the one or more processors) may be configured to implement a (trained) machine learning model that is configured to output ahealth state of the living responsive to inputting the one or more parameters 112 (e.g., a temporal change thereof). For example, the machine-learning model may be configured to classify the (e.g., temporal change of the) one or more parameters 112 regarding the health state (e.g., whether there is a heart failure, regarding a disease progression (e.g. improvement or deterioration), etc.).

[0061] According to various aspects, the radar detection method described herein may be employed to detect a plurality of radar data and to respectively determine the one or more parameters 112 as ground truth data and a machine-learning model may be trained using these ground truth data to predict the one or more parameters 112 responsive to inputting the radar data.

[0062] In the following, various examples are provided that may include one or more aspects described above with reference to the system 100 and the device 106.

[0063] Example 1 is a device including: one or more processors configured to: receive reception data representing a time series of radar reception signals reflected from one or more objects within a monitored area, the time series of radar reception signals having a sampling rate equal to or greater than 100 Hz and including at least one radar reception signal for each point in time of the time series; divide the monitored area into a plurality of three-dimensional (3D) range bins; for each range bin of the plurality of three-dimensional range bins, allocate, for each point in time of the time series, a segment of the radar reception signal to the range bin, and determine a respective amplitude value and a respective phase value using the segment of the radar reception signal; for each object of the one or more objects, determine respective one or more range bins of the plurality of three-dimensional range bins which are associated with the object; determine, using the time series of radar reception signals, whether at least one object of the one or more objects is a living being (e.g., a person); and in the case that it is determined that at least one object is a living being, determine, using a temporal change of the respective phase value of at least one range bin of the respective one or more range binsassociated with the at least one object, one or more parameters indicating a condition of a heart of the living being.

[0064] In Example 2, the subject matter of Example 1 can optionally include that the one or more parameters include a heart rate variability (representing a variation in time periods between consecutive heartbeats). Additionally, statistical parameters may be determined (e.g., calculated) over different time periods. For instance, the standard deviation of the calculated time periods may be determined over a 24-hour period. Furthermore, a spectral analysis of the heart rate variability may be performed, enabling the calculation of peak and sum values for different frequency bands. This spectral analysis allows for the extraction of frequency-domain features of HRV, such as those in the very low frequency (VLF), low-frequency (LF) and high- frequency (HF) bands.

[0065] In Example 3, the subject matter of Example 1 or 2 can optionally include that the sampling rate of the radar reception signals is equal to or greater than 250 Hz; and wherein the one or more parameters include an average (aortic) blood pressure and / or a vascular elasticity.

[0066] In Example 4, the subject matter of any one of Examples 1 to 3 can optionally include that the sampling rate of the radar reception signals is equal to or greater than 500 Hz; and wherein the one or more parameters include a ratio between a (aortic) systolic period of the heart and a (aortic) diastolic period of the heart.

[0067] In Example 5, the subject matter of any one of Examples 1 to 4 can optionally include that the sampling rate of the radar reception signals is equal to or greater than 1000 Hz; and wherein the one or more parameters include a (aortic)systolic blood pressure of the living being and / or a (aortic) diastolic blood pressure of the living being.

[0068] In Example 6, the subject matter of Example 5 can optionally include that the one or more parameters include the systolic blood pressure and the diastolic blood pressure; wherein the respective one or more range bins associated with the at least one object include at least one first range bin associated with a first position on (a body of) the living being and at least onesecond range bin associated with a second position on (the body of) the living different from the first position; and wherein the one or more processors are configured to determine the systolic blood pressure and the diastolic blood pressure using a temporal shift between the phase values of the at least one first range bin and the phase values of the at least one second range bin and using a positional difference between the first position and the second position.

[0069] In Example 7, the subject matter of any one of Examples 1 to 6 can optionally include that the radar reception signals have a frequency equal to or greater than 60 GHz.

[0070] In Example 8, the subject matter of any one of Examples 1 to 7 can optionally include that the one or more processors are configured to determine, using a (trained) machinelearning model, a health state of the living being by inputting the one or more parameters into the (trained) machine-learning model.

[0071] In Example 9, the subject matter of any one of Examples 1 to 8 can optionally include that the one or more processors are configured to determine the one or more parameters using a machine-learning model.

[0072] Example 10 is a system including: the device according any one of Examples 1 to 9; and one or more radio-frequency (RF) transceivers configured to transmit radar transmit signals in direction of the one or more objects and to receive the radar reception signals being a version of the radar transmit signals after reflection at the one or more objects.

[0073] In Example 11, the subject matter of Example 10 can optionally include that the one or more radio-frequency transceivers are configured to carry out beamforming when receiving the radar reception signal.

[0074] In Example 12, the subject matter of Example 10 or 11 can optionally include that device includes the one or more radio-frequency transceivers.

[0075] A machine-learning model, as described herein, may be or may include, for example, a reinforcement learning model (e.g., employing Q-leaming, temporal difference(TD), deep adversarial networks, etc.) and / or a classification model (e.g., a linear classifier (e.g.,logistic regression or naive Bayes classifier), a support vector machine, a decision tree, a boosted tree, a random forest, a neural network, or a nearest neighbor model). A neural network may be any kind of neural network, such as a convolutional neural network, an autoencoder network, a variational autoencoder network, a sparse autoencoder network, a recurrent neural network, a deconvolutional network, a generative adversarial network, a forward-thinking neural network, a sum-product neural network, a transformer model, etc.

[0076] The term “processor” as used herein may be understood as any kind of technological entity that allows handling of data. The data may be handled according to one or more specific functions that the processor may execute. Further, a processor as used herein may be understood as any kind of circuit, e.g., any kind of analog or digital circuit. A processor may thus be or include an analog circuit, digital circuit, mixed-signal circuit, logic circuit (e.g., a hard-wired logic circuit or a programmable logic circuit), microprocessor (for example a Complex Instruction Set Computer (CISC) processor or a Reduced Instruction Set Computer (RISC) processor), Central Processing Unit (CPU), Graphics Processing Unit (GPU), Digital Signal Processor (DSP), Field Programmable Gate Array (FPGA), integrated circuit, Application Specific Integrated Circuit (ASIC), etc., or any combination thereof. A “processor” may also be a logic-implementing entity executing software, for example any kind of computer program, for example a computer program using a virtual machine code such as for example Java. A “processor” as used herein may also include any kind of cloud-based processing system that allows handling of data in a distributed manner, e.g. with a plurality of logic-implementing entities communicatively coupled with one another (e.g. over the internet) and each assigned to handling the data or part of the data. By way of illustration, an application running on a server and the server can also be a “processor”. Any other kind of implementation of the respective functions, which will be described below in further detail, may also be understood as a processor. It is understood that any two (or more) of the processors detailed herein may be realized as a single entity with equivalent functionality or the like, and conversely that anysingle processor detailed herein may be realized as two (or more) separate entities with equivalent functionality or the like.

[0077] The term “memory” as used herein may be understood as a computer-readable medium (e.g., a non-transitory computer-readable medium), in which data or information can be stored for retrieval. References to “memory” included herein may thus be understood as referring to volatile or non-volatile memory, including random access memory (RAM), readonly memory (ROM), flash memory, solid-state storage, magnetic tape, hard disk drive, optical drive, among others, or any combination thereof. Furthermore, it is appreciated that registers, shift registers, processor registers, data buffers, among others, are also embraced herein by the term memory. It is also appreciated that a single component referred to as “memory” or “a memory” may be composed of more than one different type of memory, and thus may refer to a collective component including one or more types of memory. It is readily understood that any single memory component may be separated into multiple collectively equivalent memory components, and vice versa. Furthermore, while memory may be depicted as separate from one or more other components (such as in the drawings), it is understood that memory may be integrated within another component, such as on a common integrated chip.

[0078] In general, the term “time series” or “time-series” may refer to a sequence of data points over a time interval. Herein, a time-series may refer to data having a respective data point for each time interval of a plurality of consecutive time intervals. The time intervals of the plurality of consecutive time intervals may be equally spaced. Hence, each time interval of the plurality of consecutive time intervals may have a same length. It is understood that consecutive time intervals are characterized in that there is no gap in time between a time interval and an immediately subsequent time interval.

[0079] As used herein, the term blood pressure may also be referred to as aortic blood pressure, aortic pressure, central aortic pressure, and / or central aortic blood pressure.

[0080] While the invention has been particularly shown and described with reference to specific aspects, it should be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. The scope of the invention is thus indicated by the appended claims and all changes, which come within the meaning and range of equivalency of the claims, are therefore intended to be embraced.

Claims

ClaimsWhat is claimed is:

1. A device (106), comprising: one or more processors configured to:• receive reception data (104) representing a time series of radar reception signals reflected from one or more objects within a monitored area, the time series of radar reception signals having a sampling rate equal to or greater than 100 Hz and comprising at least one radar reception signal for each point in time of the time series;• divide the monitored area into a plurality of three-dimensional range bins;• for each range bin of the plurality of three-dimensional range bins (108), o allocate, for each point in time of the time series, a segment of the radar reception signal to the range bin, and o determine a respective amplitude value and a respective phase value using the segment of the radar reception signal;• for each object of the one or more objects (110), determine respective one or more range bins of the plurality of three-dimensional range bins which are associated with the object;• determine, using the time series of radar reception signals, whether at least one object of the one or more objects is a living being; and• in the case that it is determined that at least one object is a living being, determine, using a temporal change of the respective phase value of at least one range bin of the respective one or more range bins associated with the at least one object, one or more parameters (112) indicating a condition of a heart of the living being.

2. The device (106) according to claim 1, wherein the one or more parameters (112) include a heart rate variability.

3. The device (106) according to claim 1 or 2, wherein the sampling rate of the radar reception signals is equal to or greater than250 Hz; and wherein the one or more parameters (112) include an average blood pressure and / or a vascular elasticity.

4. The device (106) according to any one of claims 1 to 3, wherein the sampling rate of the radar reception signals is equal to or greater than500 Hz; and wherein the one or more parameters (112) include a ratio between a systolic period of the heart and a diastolic period of the heart.

5. The device (106) according to any one of claims 1 to 4, wherein the sampling rate of the radar reception signals is equal to or greater than1000 Hz; and wherein the one or more parameters (112) comprise a systolic blood pressure of the living being and / or a diastolic blood pressure of the living being.

6. The device (106) according to claim 5, wherein the one or more parameters (112) comprise the systolic blood pressure and the diastolic blood pressure; wherein the respective one or more range bins associated with the at least one object comprise at least one first range bin associated with a first position on the living being and at least one second range bin associated with a second position on the living different from the first position; and wherein the one or more processors are configured to determine the systolic blood pressure and the diastolic blood pressure using a temporal shift between the phase values of the at least one first range bin and the phase values of the at least one second range bin and using a positional difference between the first position and the second position.

7. The device (106) according to any one of claims 1 to 6, wherein the radar reception signals have a frequency equal to or greater than 60 GHz.

8. The device (106) according to any one of claims 1 to 7, wherein the one or more processors are configured to determine, using a machinelearning model, a health state of the living being by inputting the one or more parameters (112) into the machine-learning model.

9. The device (106) according to any one of claims 1 to 8, wherein the one or more processors are configured to determine the one or more parameters (112) using a machine-learning model.

10. A system (100), comprising:• the device (106) according any one of claims 1 to 9; and• one or more radio-frequency transceivers (102) configured to transmit radar transmit signals in direction of the one or more objects and to receive the radar reception signals being a version of the radar transmit signals after reflection at the one or more objects.

11. The system (100) according to claim 10, wherein the one or more radio-frequency transceivers (102) are configured to carry out beamforming when receiving the radar reception signal.

12. The system (100) according to claim 10 or 11, wherein device (106) comprises the one or more radio-frequency transceivers (102).

Images