Blood pressure determination using millimeter-wave radar

Abstract

System and method for contactless blood pressure determination. During a training phase, for each of a plurality of reference subjects, a millimeter-wave reflection radar signal reflected from a respective reference subject is received and sampled, a signal portion at subject range is extracted and segmented into segments of selected segment duration, a training dataset of training samples from plurality of reference subjects is formed, each training sample including respective segment labeled with reference blood pressure measurement for segment duration obtained using blood pressure detector, and machine learning process is applied to training dataset to generate estimation model. During a monitoring phase, a millimeter-wave reflection radar signal reflected from a monitored subject is received and sampled, a signal portion at subject range is extracted and segmented into segments of selected segment duration, and estimation model is applied to monitoring samples of segments to determine a blood pressure estimate of monitored subject.

Description

[0001] BLOOD PRESSURE DETERMINATION USING MILLIMETER-WAVE RADAR

[0002] TECHNICAL FIELD

[0003] The present disclosure generally relates to the fields of high frequency radar sensors, temporal signal processing, machine learning, and vital signs monitoring.

[0004] BACKGROUND

[0005] There are several vital signs that are commonly monitored to provide a general assessment of physical health. These vital signs indicate the status of critical bodily functions and generally include body temperature, heart rate, respiration rate, and blood pressure. Blood pressure is a measure of the pressure or force exerted by circulating blood on the inner walls of the blood vessels, particularly the arteries. This pressure is primarily produced by the contraction of the heart. Blood pressure is measured using two values: “systolic pressure”, a higher number, indicating arterial pressure when the heart contracts and is pumping blood throughout the body, and “diastolic pressure”, a lower number, indicating arterial pressure when the heart is resting between contractions. A blood pressure measurement is typically expressed in terms of the systolic pressure over the diastolic pressure and is measured in units of millimeters of mercury (mmHg) or in kilopascals (kPa), for example “120 / 70 mmHg”.

[0006] Blood pressure normally fluctuates over the course of a day and may be affected by various factors, such as: age, weight, body position, physical exertion, stress, nutrition, and medications. High blood pressure, which may be categorized as hypertension, increases the heart workload and the development of unhealthy tissue growth within the arterial walls (atheroma) that can enlarge and gradually weaken the heart muscle. Elevated blood pressure is considered a high-risk factor for cardiovascular diseases, and can lead to serious complications, including strokes; heart attacks; diabetes; transient ischemic attacks; aneurysms; peripheral artery disease; and chronic kidney failure. Low blood pressure, referred to as hypotension, can also lead to adverse effects, such as dizziness, fatigue, cardiac arrhythmias, and in severe cases may result in circulatory shock.

[0007] Blood pressure measurements may be obtained in various ways. Non-invasive measurements are typically obtained using a monitoring apparatus known as a “sphygmomanometer”, which includes an inflatable cuff, and a pressure gauge or manometer. Common forms of non-invasive measurements are the auscultatory and oscillometric techniques.

[0008] To perform auscultatory measurements, the cuff is placed around the upper arm and inflated until blood flow through the brachial artery is occluded. The air in the cuff is then gradually released. When the air pressure in the cuff falls below the systolic blood pressure in the brachial artery, blood will start to flow again through the artery. This is accompanied by a pounding sound that can be heard using a stethoscope. The pressure gauge reading at which this pounding sound begins is recorded as the systolic blood pressure. As the cuff pressure decreases further, the sounds gradually diminish and disappears completely when it falls below the diastolic blood pressure in the brachial artery as the blood flow is no longer restricted. The pressure gauge reading at this point is recorded as the diastolic blood pressure.

[0009] The oscillometric method is similar to the auscultatory but utilizes an electronic pressure transducer to detect oscillations and to automate inflation and deflation of the cuff. The cuff is initially inflated above the systolic pressure and then gradually reduced to below the diastolic pressure. The cyclic expansion and contraction of the brachial artery by the restricted blood flow causes a corresponding cyclic variation of the cuff pressure, forming an oscillation waveform over the deflation period. The oscillations in cuff pressure increase in amplitude to a maximum, which represents the “mean arterial pressure”, and then decrease in amplitude with further deflation. Various algorithms can be applied to the oscillation envelope function to extract estimates of systolic and diastolic pressure.

[0010] Pulse palpation can be used to obtain a rough estimate of minimum systolic pressure. Other non-invasive measurement techniques may be based on ultrasound, photoplethysmography (PPG), and pulse wave velocity (PWV). Invasive or direct measurements of blood pressure are less common and typically only performed for high-risk patients by trained staff. Invasive measurements are obtained by percutaneous insertion of an arterial catheter connected to a pressure transducer through a sterile fluid-filled tubing. Such invasive procedures are associated with patient discomfort, as well as risk of infections and potential complications from arterial puncturing (such as ischemia, hematoma or thrombosis), and a danger of severe bleeding following inadvertent disconnection. Invasive procedures also require technical expertise and non-reusable equipment. Invasive measurements may further be subject to errors resulting from improper catheter positioning or transducer malfunctioning or miscalibration.

[0011] Non-invasive measurements are generally quicker and less unpleasant than invasive procedures, involve fewer complications and require minimal training and moving parts. However, non-invasive measurements may deliver lower accuracy and inconsistent results depending on the method used. In particular, results may be less reliable under certain conditions, such as at pressure extremes, or for patients characterized by arrhythmia or peripheral vascular disease. Moreover, continuous monitoring is not possible. Non-invasive measurements are also subject to errors from incorrect application (e.g., improper cuff size or overly rapid deflation), interference with measurements (e.g., due to patient movement), or calculation errors. For example, oscillometric measurements may sometimes overestimate values in hypotension and underestimate in hypertension, while auscultatory measurements may underestimate blood pressure in hypotension and may be unable to detect blood pressure in low cardiac output states. The reliability of both methods is highly dependent on selecting the proper sized cuff (width and length) for the arm size of the subject.

[0012] Blood pressure detection generally entails devices or components that need to be in contact with the subject, such as coupled to or worn on the body, in order to detect variations in the volume of arterial blood flow. There are several wearable devices that incorporate blood pressure detection, such as smartwatches that can be worn on the wrist. The accuracy of wrist worn devices may be reliant on keeping the hand level with the heart. Other digital monitors may be applied on the finger or upper arm. However, such wearable devices are subject to inaccurate readings, particularly in subjects with atypical cardiac rhythm issues or arteriosclerosis. Furthermore, optical based sensors require a direct line-of-sight to the body and clear visibility, and generally cannot function or provide degraded results under poor visibility conditions, or through obstructions or occlusions such as clothing or blanketing. Some blood pressure measurement devices require a calibration process prior to use, such as to calibrate for a particular subject and / or measurement location. Such calibrations can be exceedingly cumbersome and time-consuming, particularly when applied to multiple subjects, or in changing settings that may not necessarily be known in advance.

[0013] SUMMARY

[0014] In accordance with one aspect of the present disclosure, there is thus provided a method for contactless blood pressure determination. The method includes the steps of receiving a millimeter-wave reflection radar signal reflected from at least one monitored subject; sampling the reflection radar signal and extracting a signal portion at a range of the monitored subject; segmenting the extracted signal portion into a plurality of segments, each having a selected segment duration, and applying a machine learning estimation model to monitoring samples of the segments, to determine a blood pressure estimate of the monitored subject. The machine learning estimation model may be generated during a training phase including, for each of a plurality of reference subjects, receiving a millimeter-wave reflection radar signal reflected from a respective reference subject; sampling the reflection radar signal and extracting a signal portion at a range of the reference subject; segmenting the extracted signal portion into a plurality of segments, each having a selected segment duration; forming a training dataset including training samples obtained from a plurality of reference subjects, each training sample including a respective segment labeled with a reference blood pressure measurement for the segment duration, the reference blood pressure measurement obtained using a blood pressure detector; and applying at least one machine learning process to the training dataset to generate the estimation model. The blood pressure estimate may include: a systolic blood pressure; a diastolic blood pressure; a mean arterial pressure; a blood pressure categorization; and / or a deviation from a baseline blood pressure. The method may further include the step of calculating at least one supplementary blood pressure parameter or statistic from the blood pressure. The method may further include the step of processing the reflection radar signal by applying a filter array. The extracted signal portion may include an in-phase (I) component and a quadrature (Q) component. The radar signal may be obtained using a remote non-invasive radar device including: at least one radar transmitter, configured to transmit a radar signal to a body tissue of the subject; and at least one radar receiver, configured to receive a reflection of the transmitted radar signal reflected from the body tissue of the subject. The radar signal may be a frequency- modulated continuous-wave (FMCW) radar signal. The signal portion may be sampled at a sampling rate of 500 Hz. The segment duration may be in the range of 5 to 30 seconds. The method may further include the step of issuing a notification relating to a determined blood pressure estimate.

[0015] In accordance with another aspect of the present disclosure, there is thus provided a system for contactless blood pressure determination. The system includes a radar device, configured to receive a millimeter-wave reflection radar signal reflected from at least one monitored subject. The system includes a processor, configured to sample the reflection radar signal and extract a signal portion at a range of the monitored subject; configured to segment the extracted signal portion into a plurality of segments, each having a selected segment duration, and configured to apply a machine learning estimation model to monitoring samples of the segments, to determine a blood pressure estimate of the monitored subject. The machine learning estimation model may be generated during a training phase including, for each of a plurality of reference subjects, the radar device configured to receive a millimeter-wave reflection radar signal reflected from a respective reference subject, and the processor configured to sample the reflection radar signal and extract a signal portion at a range of the reference subject; configured to segment the extracted signal portion into a plurality of segments, each having a selected segment duration; configured to form a training dataset including training samples obtained from a plurality of reference subjects, each training sample including a respective segment labeled with a reference blood pressure measurement for the segment duration, the reference blood pressure measurement obtained using a blood pressure detector; and configured to apply at least one machine learning process to the training dataset to generate the estimation model. The blood pressure estimate may include: a systolic blood pressure; a diastolic blood pressure; a mean arterial pressure; a blood pressure categorization; and / or a deviation from a baseline blood pressure. The processor may be further configured to calculate at least one supplementary blood pressure parameter or statistic from the blood pressure. The processor may be further configured to process the reflection radar signal by applying a filter array. The extracted signal portion may include an in-phase (I) component and a quadrature (Q) component. The radar device may include a remote non-invasive radar device including: at least one radar transmitter, configured to transmit a radar signal to a body tissue of the subject; and at least one radar receiver, configured to receive a reflection of the transmitted radar signal reflected from the body tissue of the subject. The radar signal may be a frequency-modulated continuous-wave (FMCW) radar signal. The signal portion may be sampled at a sampling rate of 500 Hz. The segment duration may be in the range of 5 to 30 seconds. The system may further include a notification unit, configured to issue a notification relating to a determined blood pressure estimate.

[0016] BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present disclosure will be understood and appreciated more fully from the following detailed description taken in conjunction with the drawings in which:

[0018] Figure 1 is a schematic illustration of a contactless blood pressure determination system, constructed and operative in accordance with an embodiment of the present disclosure;

[0019] Figure 2 is a schematic illustration of the radar device of the system of Figure 1 receiving reflections from a subject on a chair, constructed and operative in accordance with an embodiment of the present disclosure;

[0020] Figure 3 is a block diagram of a contactless blood pressure determination method, operative in accordance with an embodiment of the present disclosure;

[0021] Figure 4 is a flow diagram of a model training phase of a contactless blood pressure determination method, operative in accordance with an embodiment of the present disclosure; and

[0022] Figure 5 is a flow diagram of a subject monitoring phase of a contactless blood pressure determination method, operative in accordance with an embodiment of the present disclosure.

[0023] DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] The disclosure embodiments may overcome the disadvantages of the prior art by providing a method and system for measuring blood pressure of a subject, with a high degree of accuracy and in a contact free manner, and without requiring a time-consuming calibration process.

[0025] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the disclosed subject matter belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the specification and claims and should not be interpreted in an idealized or overly formal sense unless expressly so defined herein. Well-known functions or constructions may not be described in detail for brevity and / or clarity.

[0026] It will be understood that, although the terms first, second, etc., may be used herein to describe various elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited by these terms. Rather, these terms are only used to distinguish one element, component, region, layer and / or section, from another element, component, region, layer and / or section.

[0027] It will be understood that when an element is referred to as being “on”, “attached” to, “operatively coupled” to, “operatively linked” to, “operatively engaged” with, “connected” to, “coupled” with, “contacting”, “added to, another element, it can be directly on, attached to, connected to, operatively coupled to, operatively engaged with, coupled with, added to, and / or contacting the other element or intervening elements can also be present. In contrast, when an element is referred to as being “directly contacting” another element or “directly added” to another element, there are no intervening elements and / or steps present.

[0028] Whenever the term “about” or “approximately” is used, it is meant to refer to a measurable value such as an amount, a temporal duration, and the like, and is meant to encompass variations (e.g., ±20%, ±10%, ±5%, ±1 %, ±0.1 %) from the specified value, as such variations are appropriate to perform the disclosed methods.

[0029] Certain features of the disclosed subject matter, which are, for clarity, described in the context of separate embodiments, may also be provided in combination in a single embodiment. Conversely, various features of the disclosed subject matter, which are, for brevity, described in the context of a single embodiment, may also be provided separately or in any suitable sub-combination or as suitable in any other described embodiment of the disclosure. Certain features described in the context of various embodiments are not to be considered essential features of those embodiments, unless the embodiment is inoperative without those elements.

[0030] Throughout this application, various embodiments of the present disclosure may be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the disclosed embodiments. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range, regardless of the breadth of the range. Whenever a numerical range is indicated herein, it is meant to include any cited numeral (fractional or integral) within the indicated range.

[0031] Whenever terms “plurality” and “a plurality” are used it is meant to include, for example, “multiple” or “two or more”. The terms “plurality” or “a plurality” may be used throughout the specification to describe two or more components, devices, elements, units, parameters, or the like. The term set when used herein may include one or more items. Unless explicitly stated, the method embodiments described herein are not constrained to a particular order or sequence. Additionally, some of the described method embodiments or elements thereof can occur or be performed simultaneously, at the same point in time, or concurrently.

[0032] The term “repeatedly” as used herein should be broadly construed to include any one or more of: “continuously”, “periodic repetition” and “nonperiodic repetition”, where periodic repetition is characterized by constant length intervals between repetitions and non-periodic repetition is characterized by variable length intervals between repetitions.

[0033] The terms “user” and “operator” are used interchangeably herein to refer to any individual person or group of persons using, operating or performing at least one aspect of a method or system of a disclosed embodiment, such as a person performing a blood pressure measurement of a subject.

[0034] The term “subject” is used herein to refer to an individual upon which a method or system of a disclosed embodiment is performed, such as a person for whom a blood pressure measurement is determined. The subject may be any living entity, such as a person, human or animal, characterized with a functioning heartbeat associated with a cardiac cycle of the heart.

[0035] Reference is now made to Figure 1 , which is a schematic illustration of a contactless blood pressure determination system, generally referenced 110, constructed and operative in accordance with an embodiment of the present disclosure. System 110 includes a radar device 112, a processor 114, a blood pressure detector (BPD) 116, a user interface 117, a database 118, and a notification unit 119. Processor 114 is communicatively coupled with radar device 112, with BPD 116, with user interface 117, with database 118, and with notification unit 119.

[0036] Radar device 112 is configured to transmit a radar signal 122 toward an intended subject 120 in a monitoring area. Radar device 112 is further configured to receive a reflection radar signal 124 respective of the transmitted radar signal 122, such as a reflection signal reflected from a body part of subject 120, such as a chest area. In one example, the monitoring area includes a supporting surface for supporting a subject 120, such as a chair or a bed. Reference is further made to Figure 2, which is a schematic illustration of radar device 112 of system 110 receiving reflections from a subject 120 on a chair 130, constructed and operative in accordance with an embodiment of the present disclosure. Figure 2 shows system 110 applied on a subject 120 sitting on chair 130, although may alternatively be applied with the subject supported by a different type of supporting surface and / or with the subject in a different body position or alignment, such as standing, lying down (recumbent), or reclined. Radar device 112 may be positioned in the vicinity of chair 130, such as mounted or coupled to a structure above chair 130 (e.g., a wall, a ceiling, a post) and directed to a supporting surface of chair 130, such that a field of view (FOV) of radar device 112 encompasses at least a portion of chair 130 where subject 120 may be present. Radar device 112 may be positioned at a predetermined distance from subject 120, such as within a few meters.

[0037] Radar device 112 includes at least one radar transmitter and at least one radar receiver. Each radar transmitter and radar receiver may be embodied by one or more transmitting / receiving elements, such as an individual radar antenna or an array of radar antennas, for example a phased array radar, such as a multiple-input multiple-output (MIMO) radar. The transmitted radar signal 122 may be at a sufficiently high frequency to ensure that the signal is reflected and not absorbed by the body tissue, for example in the millimeter wave (MMW) frequency band (corresponding to EHF radio frequencies). According to an embodiment, the transmitted and reflected radar signals 122, 124 are in the Terahertz (THz) frequency band, where the term “Terahertz (THz)” as used herein encompasses Terahertz and sub-Terahertz radiation corresponding to submillimeter and millimeter wave radiation, such as electromagnetic waves within the frequency band between about 0.03 to 3 THz, corresponding to radiation wavelengths between about 10 mm to 0.1 mm. In one example, transmitted and reflected radar signals 122, 124 is at a frequency above 100 GHz, such as approximately 120 GHz (or 0.12 THz).

[0038] Radar device 112 may be as described for example in PCT application publication WO2018 / 167777A1 to Neteera Technologies, entitled “Method and device for non-contact sensing of vital signs and diagnostic signals by electromagnetic waves in the sub terahertz band”, and PCT application publication W02020 / 012455A1 to Neteera Technologies, entitled “A sub-THz and THz system for physiological parameters detection and method thereof”. It is noted that radar device 112 operates in a contactless manner, which transmits and receives radar signals remotely without requiring a device component to be in direct physical contact with subject 120 or to be worn or attached to subject 120. It is further noted that radar device 112 may transit and / or receive a reflected radar signal from any direction of subject 120, such as from a front or back direction or from a non-orthogonal angle relative to subject 120. Moreover, radar device 112 may transmit and receive a reflected radar signal in low light or poor visibility conditions, as well as through certain obstructions or material barriers covering the subject body, where the radar signal may penetrate through clothing worn by subject 120, or a fabric or other material of a supporting surface (e.g., a chair 130) on which subject 120 is positioned.

[0039] Blood pressure detector (BPD) 116 obtains blood pressure measurements during a learning or training phase of system 110. The blood pressure measurements may include readings relating to one or more blood pressure parameters, including but not limited to: systolic blood pressure (SBP); diastolic blood pressure (DBP); and mean arterial pressure (MAP). BPD 116 may be embodied by one or more devices configured to detect at least one blood pressure measurement, using any suitable blood pressure measurement technique, including but not limited to: non-invasive measurements; invasive measurements; auscultatory measurements; oscillometric measurements; and measurement techniques based on ultrasound, photoplethysmography (PPG), optical detection; and / or pulse wave velocity (PWV) principles. BPD 116 may include contacting or wearable sensors or measurement devices.

[0040] Processor 114 receives information or instructions from other components of system 110 and performs required data processing. For example, processor 114 receives and processes reflected radar signals 124 obtained by radar device 112 to generate a machine learning model during a data acquisition or modelling phase, and to determine a blood pressure measurement during a subject monitoring phase, as will be elaborated upon further hereinbelow.

[0041] Database 118 stores relevant information to be retrieved and processed by processor 114, such as radar signal data and associated information. Database 118 may be represented by one or more local servers or by remote and / or distributed servers, such as in a cloud storage platform.

[0042] User interface 117 allows a user to receive information and to control parameters or settings associated with system 110. For example, user interface 117 may include a display screen configured to present visual content. User interface 117 may include a cursor and / or a touch-screen menu interface, such as a graphical user interface, configured to enable manually entering instructions or data. User interface 117 may also include peripheral communication devices configured to provide audible communication, such as a microphone and an audio speaker, and may include voice recognition capabilities to enable the user to enter instructions or data by means of speech commands.

[0043] Notification unit 119 is configured to provide alerts or notifications relating to the operation of system 110. For example, notification unit 119 may issue an alert when a determined blood pressure measurement exceeds a threshold value. Notification unit 119 may utilize various notification modes, such as visual indications (e.g., displaying text, markings and / or symbols, changing colors of graphical content) and / or audible indications (e.g., alarms, beeps, buzzers, bells, ringtones). Accordingly, notification unit 119 may include one or more devices or instruments configured for audio communication (e.g., an audio speaker, a speech generator) and / or visual communication (e.g., a display screen; a graphical user interface; flashing lights). Notification unit 119 may optionally be integrated with user interface 117 or may be a separate unit. Notification unit 119 may be configured to provide alerts or notifications to a remote location, such as by transmitting an indication to at least one of: an operator computer device (e.g., via a dedicated application); a cloud service, a remote internet application, a webbased communication network; a supervisor or central monitoring station (e.g., associated with a health care facility), and the like. A plurality of notification units 119 may be employed for providing alerts at different locations.

[0044] Information may be conveyed between the components of system 110 over any suitable data communication channel or network, using any type of channel or network model and any data transmission protocol (e.g., wired, wireless, radio, WiFi, Bluetooth, and the like). For example, system 110 may store, manage and / or process data using a cloud computing model, and the components of system 110 may communicate with one another and be remotely monitored or controlled over the Internet, such as via an Internet of Things (loT) network. The components and devices of system 110 may be based in hardware, software, or combinations thereof. It is appreciated that the functionality associated with each of the devices or components of system 110 may be distributed among multiple devices or components, which may reside at a single location or at multiple locations. For example, the functionality associated with processor 114 may be distributed between a single processing unit or multiple processing units (e.g., a dedicated machine learning processor for the data modeling phase). Processor 114 may be part of a server or a remote computer system accessible over a communications medium or network, such as a cloud computing platform. Processor 114 may also be integrated with other components of system 110, such as incorporated with radar device 112. System 110 may optionally include and / or be associated with additional components not shown in Figure 1 , for enabling the implementation of the disclosed subject matter.

[0045] The operation of system 110 will now be described in general terms, followed by specific examples. Reference is now made to Figure 3, which is a block diagram of a contactless blood pressure determination method, operative in accordance with an embodiment of the present disclosure. In step 152, a radar signal reflected from a subject is received. Referring to Figures 1 and 2, radar device 112 transmits a coherent radar signal 122 in the THz frequency band, such as a continuous-wave (CW) or frequency-modulated continuous wave (FMCW) radar signal, to a subject 120. Radar device 112 may receive a corresponding reflected radar signal 124 from a body part of subject 120, where reflection signal 124 contains information relating to micro displacements in the skin associated with the cardiac cycle and pulmonary activity of subject 120. Transmitted radar signal 122 may be directed to a selected body area of subject, such as a chest or back, and may be directed to an anterior (frontal) body portion or a posterior (rear) body portion of subject 120.

[0046] In step 153, the reflected signal is sampled with in-phase and quadrature components and a signal portion at the range of the subject is extracted. Referring to Figures 1 and 2, processor 114 receives and filters reflected radar signal 124 based on signal phase differences that correlate with distance, to isolate signal components at the range at which subject 120 is located. In particular, processor 114 receives the reflected radar signal 124, obtained at a selected sampling rate and recorded with two channels consisting of an in-phase (I) component and a quadrature (Q) component. For example, reflected radar signal 124 may be sampled at a rate of 500 Hz, such that 500 times per second a complex valued signal composed of in-phase (I) and quadrature (Q) data is collected. Processor 114 may determine the distance traversed by the propagating radar signal, to extracts a signal portion corresponding to reflections from subject 120 and removing noise and irrelevant signal components at other ranges. For example, if transmitted and reflected radar signals 122, 124 are FMCW radar signals, then processor 114 may apply a fast Fourier transform (FFT) to extract the signal portion at a preselected range bin corresponding to the subject range. A subject range determination may be implemented, for example, in accordance with methods described in PCT application publication WO2013 / 275865A1 to Neteera Technologies, entitled “Radar-based range determination and validation”.

[0047] In step 155, the extracted signal portion is processed. Referring to Figures 1 and 2, processor 114 applies processing operations to the extracted portion of reflected signal 124 at the subject range. The processing operations may include the application of a linear mapping to produce an encoding in the frequency domain. For example, the extracted signal portion with in-phase and quadrature components (i.e., “IQ signal”) may be subject to a plurality of time invariant linear filters, such as bandpass filters, which may be obtained from a predefined filter bank, to produce a set of filter responses. Alternatively, the linear mapping may involve a Fourier transform, or a convolution operator, rather than linear filtering. The processing may further include one or more non-linear mapping operations, such as: rectification (i.e., obtaining the absolute value of the (complex) signal), followed by exponentiation (e.g., squaring), followed by gain control (i.e., division by the average energy in the signal over a brief time period).

[0048] In step 156, the processed signal is segmented. Referring to Figures 1 and 2, processor 114 divides the processed signal into discrete temporal segments of a selected time duration “N”, to produce vectors of a selected length, where the segment duration is an adjustable parameter. For example, the segment duration may be in the range of 5 to 30 seconds, such as approximately 10 seconds. If the frequency of the IQ signal is 500Hz, then a segmentation of 10 seconds duration produces a vector data set of length 5000 (10 seconds X 500 Hz = 5000). It is noted that the signal processing (step 155) may be applied before or after the segmentation (step 156), such that the linear mapping may alternatively be applied to each segment. The processing and segmentation results in multiple filter responses for a given segment, which may be represented as a matrix of size “segmentjength" x “njilters", where “segmentjength" is the length or duration of the signal segment (i.e., the number of samples), and “n_filters" is the number of filters used to create the filter responses (e.g., number of time invariant filters in the filter bank).

[0049] The aforementioned sequence is repeated for multiple reference subjects to obtain a collection of training samples (vector segments) during an initial data acquisition and model training phase 160. In step 162 of the training phase 160, training samples of reference vector segments are obtained, each vector segment having a selected segment duration and assigned with a respective label reflecting a reference blood pressure measurement. Referring to Figure 1 , system 110 obtains a large number of vector segments from respective reference subjects 120, by performing steps 152, 153, 155, 156, with each obtained vector segment representing a training sample. Each reference subject is monitored with blood pressure detector 116 which provides reference measurements of blood pressure. For each vector segment, processor 114 examines blood pressure readings of reference subject 120 obtained by BPD 116 and assigns a corresponding reference blood pressure measurement for the time duration of that segment (e.g., 10 seconds). Depending on the mode of measurement, BPD 116 may provide continuous or intermittent readings. For example, if BPD 116 employs cuff-based (auscultatory) measurements, then readings are typically obtained intermittently (in view of the required time for cuff inflation and deflation prior to each reading), in which case segments of radar signals received during the reference measurement period would get assigned a common reference measurement based on the reading at the end of the time period. Alternatively, if BPD 116 employs arterial line measurements (i.e., invasive measurements using an arterial catheter), then the readings are generally obtained with a much higher temporal resolution, in which case each segment of radar data is assigned an instantaneous reading produced by the arterial line detection. Each training sample or vector segment is assigned a primary reference measurement reflective of at least one quantitative blood pressure value of the reference subject 120 detected during the duration of the vector segment. The assigned reference measurement may include one or more blood pressure values, such as a systolic blood pressure (SBP), a diastolic blood pressure (DBP), and / or a mean arterial pressure (MAP). The reference measurement may include a blood pressure classification determined from readings of BPD 116, such as a classification selected from the categories of: a high blood pressure, a normal blood pressure, and a low blood pressure. An exemplary categorization for an adult may be as follows: “normal” (e.g., SBP=120-129; DBP=80-84); “elevated” (e.g., SBP=130-139; DBP=85-89); “hypertension” (e.g., SBP>140; DBP>90); “low” (e.g., SBP=90-119; DBP=60-79); and “hypotension” (e.g., SBP<90; DBP<60). Each training sample may optionally be assigned additional subject identification labels, such as: a name, an identification number, and / or other personal information relating to the reference subject (e.g., age, gender, location, physical attributes) to facilitate subsequent data analysis. The vector datasets and associated blood pressure labels may be stored in database 118 and used as training data to generate a machine learning model for application during a subject monitoring stage.

[0050] In step 164 of the model training phase 160, a machine learning process is applied to a collection of training samples for training a blood pressure estimation model. The training dataset, representing vector segments and associated blood pressure labels for a large number of reference subjects, is analyzed using a machine learning process, to implicitly identify different patterns and create models for estimating blood pressures of monitored subjects. The machine learning process may apply machine learning techniques to analyze the training data, in order to produce mapping functions that can be used for classifying additional instances of new datasets (vector segments) according to relevant classification criteria. The data analysis may utilize any suitable machine learning or supervised learning process or algorithm, including but not limited to: an artificial neural network (ANN) process, such as a convolutional neural network, recurrent neural network (RNN), or a deep learning algorithm; a classification or regression analysis, such as a linear regression model; a logistic regression model, or a support-vector machine (SVM) model; a decision tree learning approach, such as a random forest classifier; and / or any combination thereof. The data analysis may utilize any suitable tool or platform, such as publicly available opensource machine learning or supervised learning tools.

[0051] Processor 114 may establish classification profiles of reference subjects by assigning reference subjects into different groups or categories based on common features. For example, the training process may provide models for identifying subjects in various categories, such as based on: age, gender, location, physical attributes, BP classifications, and the like. Each category may be associated with a relative weighting metric corresponding to a confidence level pertaining to the respective category feature. The different models may then be applied to facilitate the blood pressure estimation of a new subject belonging to the relevant category or profile. The generated models may be iteratively updated and improved based on new information, such as accounting for the accuracy of subsequent blood pressure estimations of monitored subjects, and additional training data collected from new reference subjects. Simulations of numerous collected data and estimated blood pressure values may also be applied to enhance the reliability and accuracy of the models. The updated models may provide optimal formulas and weighting metrics for different variables or classification features. As more information and statistics are accumulated, the models can be further refined to improve their predictive capability for subject identification.

[0052] Reference is made to Figure 4, which is a flow diagram of a model training phase of a contactless blood pressure determination method, operative in accordance with an embodiment of the present disclosure. A reflection radar signal 212 is obtained from a reference subject 220 using a contactless radar device 112. Reflection radar signal 212 is sampled and filtered to extract a signal portion 214 at the subject range having in-phase (I) and quadrature (Q) components. The extracted IQ signal 214 is processed, such as by applying a filter array. The processed (filtered) signal 218 is divided into discrete temporal segments of equal duration to form a reference dataset of vector segments 222, each having a vector length corresponding to the segment duration. A blood pressure detector reading 224 is obtained from reference subject 220 using BPD 116. Each vector in reference dataset 222 is assigned a blood pressure label 226 based on BPD reading 224, representing a reference measurement of blood pressure for the segment duration of the respective vector segment. This process is repeated for a large number of reference subjects 220 to generate a collection of training samples 228, each training sample made up of a vector segment 222 and assigned blood pressure label (reference measurement) 226. The collection of training samples 228 is fed into a training model 230 that utilizes machine learning techniques to produce a blood pressure estimation model 232 that can be applied on new input datasets. The generated blood pressure estimation model 232 may be dynamically updated and subsequently applied to determine a blood pressure estimate, and supplementary blood pressure parameters and statistics, of a monitored subject during a subject monitoring phase 170. A collection of training samples 228 may include samples obtained from different sources, such as different types of BPDs 116 employing different measurement modalities, which may provide variation to augment the training process.

[0053] Reference is made to Figure 5, which is a flow diagram of a subject monitoring phase of a contactless blood pressure determination method, operative in accordance with an embodiment of the present disclosure. A dataset of monitoring samples of a monitored subject 320 is obtained according to a similar sequence described hereinabove in the context of model training phase 160. In particular, a reflection radar signal 312 is obtained from a monitored subject 320 using a contactless radar device 112. Reflection radar signal 312 is sampled and filtered to extract a signal portion 314 at the subject range having in-phase (I) and quadrature (Q) components. The extracted IQ signal 314 is processed, such as by applying a filter array. The processed (filtered) signal 318 is divided into discrete temporal segments of equal duration to form a dataset of vector segments, each having a vector length corresponding to the segment duration. The vector segments, accumulated over the selected sample duration, make up monitoring samples 324, which are fed into estimation model 232.

[0054] Referring back to Figure 3, in step 172 of the subject monitoring phase 170, monitoring samples of segments from a monitored subject are obtained, each segment having a selected segment duration. Referring to Figure 5, monitoring dataset 326, made up of a plurality of samples 324 of vector segments derived from radar reflections 312 from monitored subject 320, are collected over a given time period and fed into blood pressure estimation model 232. Vector segments of monitoring samples 324 are characterized by a vector length corresponding to the segment duration of the segments (applied when segmenting processed signal 318), corresponding to the segment duration of the reference subject segments. Monitoring samples 324 may be provided to blood pressure estimation model 232 as a continuous input stream, such that the radar signals are segmented as acquired and the segments fed serially (i.e., one by one) into model 232. Alternatively, samples 324 may be delivered in parallel to blood pressure estimation model 232 (i.e., “offline”) after waiting to acquire and segment a longer input signal. The particular segments of monitoring samples 324 that are processed may be dynamically selected. It is noted that consecutive segments may overlap, whereby consecutive segments of a certain duration (e.g., 10 second segments) may overlap by a certain amount (e.g., 9 seconds).

[0055] In step 174 of the subject monitoring phase 170, a blood pressure estimation model is applied to the monitoring samples to obtain a blood pressure estimate. Referring to Figure 5, blood pressure estimation model 232 applies machine learning processes to the monitoring samples 324 to determine an estimation of blood pressure for the duration of the segments, based on earlier training of model 232 using labeled segments having the same duration. Model 232 provides an estimate of blood pressure of monitored subject 320 for each sample segment 324. The blood pressure estimate may include one or more blood pressure parameters, such as a systolic blood pressure (SBP), a diastolic blood pressure (DBP), and / or a mean arterial pressure (MAP). The blood pressure estimate may include a blood pressure category or classification, such as in relation to predefined thresholds or selected criteria. An exemplary categorization for an adult may be as follows: “normal” (e.g., SBP=120-129; DBP=80-84); “elevated” (e.g., SBP=130-139; DBP=85-89); “hypertension” (e.g., SBP>140; DBP>90); “low” (e.g., SBP=90-119; DBP=60-79); “hypotension” (e.g., SBP<90; DBP<60). For example, model 232 may classify subject as belonging to a category of “high BP” (e.g., elevated or hypertension ranges), “medium BP” (e.g., normal range), or “low BP” (e.g., low or hypotension ranges). Model 232 may indicate a deviation from a baseline blood pressure value or range, such as whether monitored subject 320 exceeds or falls below the baseline. If monitored subject 320 is determined to have moved or is absent from the monitoring location for a certain duration during the monitoring period, such as based on an indication of subject occupancy (e.g., from a motion or proximity sensor), then a blood pressure estimate may not be determined or may be voided during the subject absence, and a notification may be provided accordingly. In step 178 of the subject monitoring phase 170, supplementary blood pressure parameters or statistics of the monitored subject is determined. Referring to Figures 1 and 5, processor 114 calculates one or more parameters or statistics from the determined blood pressure estimate, such as a maximum BP, a minimum BP, an average BP over a certain duration, a standard deviation, and the like. Processor 114 may determine a blood pressure category of classification based on the blood pressure estimate according to predefined thresholds or selected criteria, such as a categorization of: “normal”; “elevated”; “hypertension”; “low”; or “hypotension”. Alternatively or additionally, a blood pressure categorization of monitored subject 320 may be part of a blood pressure estimate provided by estimation model 232. Processor 114 may determine whether the blood pressure estimate deviates from a baseline BP value or range, such as if the estimate exceeds an upper baseline threshold or falls below a lower baseline threshold. It is noted that if a particular segment contains an absence interval that exceeds a certain threshold (e.g., a high percentage of seconds in which the subject is not present), then that segment may be removed from the BP parameter calculation or categorization.

[0056] Monitored subject 320 may undergo additional monitoring in subsequent sessions, such as over a period of several days, weeks, or months, from which further blood pressure information and statistics may be determined (e.g., average BP over multiple sessions; peak BP; standard deviations; and the like). Such statistics may be reviewed and analyzed to provide focused behavioral recommendations for monitored subject 320, in order to improve physical health.

[0057] Notification unit 119 may issue a notification or alert relating to a determined blood pressure of monitored subject 320. For example, notification unit 119 may issue an alert when a blood pressure estimate or associated parameter (such as average BP) exceeds a predetermined upper threshold or falls below a predetermined lower threshold. The notification may be provided as a visual indication and / or an audible indication. An issued notification may be linked to one or more conditions, which may be configured or modified by an operator. For example, an alert may be issued if a monitored subject 320 is detected to be in an abnormal (e.g., elevated or low) blood pressure state for a predefined minimum time interval. The alert may be communicated to an external monitoring system and / or at least one operator, such as a dedicated clinician at a facility, and an action may be undertaken responsive to the alert, such as to perform a physical evaluation of subject 320. A notification or alert may be associated with a priority level reflecting a degree of importance or urgency, where different notifications or alerts may be presented differently based on their priorities. For example, a first alert having a high urgency level (e.g., reflecting an elevated or low blood pressure determination of a subject 320 with severe health complications) may be issued as a high volume or high frequency audible sound and / or a visually intense graphical symbol, whereas a second alert having a low urgency level (e.g., reflecting an elevated or low blood pressure determination of a subject 320 with minimal health issues) may be issued as a low volume or low frequency audible sound and / or a graphical symbol with low intensity.

[0058] More generally, system 110 may provide a report, such as via user interface 117, relating to determined blood pressure estimates and / or associated parameters and statistics of monitored subject 320. The report may include a visual representation of determined blood pressure information over a selected monitoring period, and optional supplementary information relating to subject 120. The report may include characteristics of determined blood pressure information, such as duration and frequency of abnormal blood pressure states (e.g., elevated, hypertension, low, hypotension) during a given time interval. The report may further include various statistics associated with the determined blood pressure information, such as historical data obtained at previous dates and times, or blood pressure parameters determined for similar types of subjects. The information and statistics may be presented over a selected duration, such as depicting dynamic changes over time, which may provide insight into the behaviour and general wellbeing of monitored subject 320. Notification unit 119 may issue a notification or alert based on deviations of blood pressure estimates relative to statistical or historical data, which may provide early warning signs of changes in a physical condition of subject 320.

[0059] The method of Figures 3, 4 and 5 may be implemented in an iterative manner, such that at least some of the steps are performed repeatedly, in order to provide for a dynamic determination of blood pressure and associated parameters, categorizations and statistics of one or more subjects in real-time. It is appreciated that the disclosed embodiments may provide a reliable and accurate determination of blood pressure of a monitored subject over time, as determined using machine learning processes. Further parameters and calculated statistics obtained over subsequent sessions can provide a more comprehensive evaluation and form the basis for providing targeted health recommendations. The disclosed system does not require components to be in direct physical contact with the monitored subject, and there is no need for coupling a sensor or other device to the subject body before a monitoring session, thus saving time and minimizing discomfort. Furthermore, the subject does not need to be directly visible to the radar device, which may operate under poor visibility or light saturation conditions. The radar signal may be reliably obtained through obstructions or occlusions, such as clothing or blankets, and from different angles in relation to the subject. Source separation techniques may be applied for measuring and identifying multiple subjects concurrently. The disclosed embodiments do not require costly equipment and has relatively few components and is relatively straightforward to operate and maintain. The BP determination of the disclosed method and system may be implemented in a wide variety of locations (e.g., rooms or facilities), and different types of subjects (e.g., regardless of age, height, weight, or other physical characteristics), without requiring a timeconsuming and cumbersome calibration process prior to each individual monitoring session depending on the type of location or subject. Moreover, a machine learning analysis may provide reliable and accurate predictive models, which can be iteratively refined to improve the BP estimation of new subjects based on new information.

[0060] The disclosed system and method can be used for various applications, ranging from home healthcare to medical diagnosis. For example, the disclosed system and method may be applied for monitoring patients to evaluate physical health in an eldercare facility.

[0061] While certain embodiments of the disclosed subject matter have been described, so as to enable one of skill in the art to practice the present disclosure, the preceding description is intended to be exemplary only. It should not be used to limit the scope of the disclosed subject matter, which should be determined by reference to the following claims.

Claims

CLAIMS1. A method for contactless blood pressure determination, the method comprising the steps of: receiving a millimeter-wave reflection radar signal reflected from at least one monitored subject; sampling the reflection radar signal and extracting a signal portion at a range of the monitored subject; segmenting the extracted signal portion into a plurality of segments, each having a selected segment duration; and applying a machine learning estimation model to monitoring samples of the segments, to determine a blood pressure estimate of the monitored subject.

2. The method of claim 1 , wherein the machine learning estimation model is generated during a training phase comprising: for each of a plurality of reference subjects, receiving a millimeter-wave reflection radar signal reflected from a respective reference subject; sampling the reflection radar signal and extracting a signal portion at a range of the reference subject; segmenting the extracted signal portion into a plurality of segments, each having a selected segment duration; forming a training dataset comprising training samples obtained from a plurality of reference subjects, each training sample comprising a respective segment labeled with a reference blood pressure measurement for the segment duration, the reference blood pressure measurement obtained using a blood pressure detector; and applying at least one machine learning process to the training dataset to generate the estimation model.

3. The method of claim 1 , wherein the blood pressure estimate is selected from the group consisting of: a systolic blood pressure;a diastolic blood pressure; a mean arterial pressure; a blood pressure categorization; and a deviation from a baseline blood pressure.

4. The method of claim 1 , further comprising the step of calculating at least one supplementary blood pressure parameter or statistic from the blood pressure estimate.

5. The method of claim 1 , further comprising the step of processing the reflection radar signal by applying a filter array.

6. The method of claim 1 , wherein the extracted signal portion comprises an in-phase (I) component and a quadrature (Q) component.

7. The method of claim 1 , wherein the radar signal is obtained using a remote non-invasive radar device comprising: at least one radar transmitter, configured to transmit a radar signal to a body tissue of the subject; and at least one radar receiver, configured to receive a reflection of the transmitted radar signal reflected from the body tissue of the subject.

8. The method of claim 1 , wherein the radar signal is a frequency-modulated continuous-wave (FMCW) radar signal.

9. The method of claim 1 , wherein the signal portion is sampled at a sampling rate of 500 Hz.

10. The method of claim 1 , wherein the segment duration is in the range of 5 to 30 seconds.

11. The method of claim 1 , further comprising the step of issuing a notification relating to a determined blood pressure estimate.

12. A system for contactless blood pressure determination, the system comprising: a radar device, configured to receive a millimeter-wave reflection radar signal reflected from at least one monitored subject; and a processor, configured to sample the reflection radar signal and extract a signal portion at a range of the subject, to segment the extracted signal portion into a plurality of segments, each having a selected segment duration, and to apply a machine learning estimation model to monitoring samples of the segments, to determine a blood pressure estimate of the monitored subject.

13. The system of claim 12, wherein the machine learning estimation model is generated during a training phase comprising: for each of a plurality of reference subjects, the radar device is configured to receive a millimeter-wave reflection radar signal reflected from a respective reference subject; and the processor is configured to sample the reflection radar signal and extract a signal portion at a range of the reference subject; to segment the extracted signal portion into a plurality of segments, each having a selected segment duration; to form a training dataset comprising training samples obtained from a plurality of reference subjects, each training sample comprising a respective segment labeled with a reference blood pressure measurement for the segment duration, the reference blood pressure measurement obtained using a blood pressure detector; and to apply at least one machine learning process to the training dataset to generate the estimation model.

14. The system of claim 12, wherein the blood pressure estimate is selected from the group consisting of: a systolic blood pressure; a diastolic blood pressure;a mean arterial pressure; a blood pressure categorization; and a deviation from a baseline blood pressure.

15. The system of claim 12, wherein the processor is further configured to calculate at least one supplementary blood pressure parameter or statistic from the blood pressure estimate.

16. The system of claim 12, wherein the processor is further configured to process the reflection radar signal by applying a filter array.

17. The system of claim 12, wherein the extracted signal portion comprises an in-phase (I) component and a quadrature (Q) component.

18. The system of claim 12, wherein the radar device comprises a remote non-invasive radar device comprising: at least one radar transmitter, configured to transmit a radar signal to a body tissue of the subject; and at least one radar receiver, configured to receive a reflection of the transmitted radar signal reflected from the body tissue of the subject.

19. The system of claim 12, wherein the radar signal is a frequency-modulated continuous-wave (FMCW) radar signal.

20. The system of claim 12, wherein the signal portion is sampled at a sampling rate of 500 Hz.21 . The system of claim 12, wherein the segment duration is in the range of 5 to 30 seconds.

22. The system of claim 12, further comprising a notification unit, configured to issue a notification relating to a determined blood pressure estimate.

Images