Bed occupancy monitoring using millimeter wave radar

Abstract

System and method for bed occupancy monitoring. Millimeter-wave reflection radar signal received from bed in monitoring area. Reflection radar signal sampled and signal portion extracted at range of subject of bed, signal portion consisting of in-phase and quadrature signal components. Median zero- crossing (MZC) derived over selected time window from whichever signal component has larger magnitude. Bed occupancy state of bed is determined based on MZC, by comparing to a middle threshold, determining an empty bed state when MZC is above middle threshold, and determining an occupied bed state when MZC below middle threshold. Transitional occupancy state is determined by comparing MZC to a low threshold and a high threshold, determining an entering bed state when MZC below low threshold for lag period, and determining an exiting bed state when MZC above high threshold and waveform characteristic detected in a signal component or MZC exceeds high threshold for sustained period.

Description

[0001] BED OCCUPANCY MONITORING USING

[0002] MILLIMETER WAVE RADAR

[0003] TECHNICAL FIELD

[0004] The present disclosure generally relates to the fields of high frequency radar sensors, temporal signal processing, machine learning, and sleep monitoring.

[0005] BACKGROUND

[0006] The occupancy status of a bed may be significant for several reasons. Certain individuals, such as those who are elderly, debilitated or suffering from an illness, may require assistance with leaving or entering a bed, such as for overcoming physical difficulties and avoiding possible injuries. In some cases, a person may be confined under medical supervision and may be permitted to leave the bed only in certain circumstances. A person may accidentally fall out of bed and require assistance but be physically incapable of calling for help. On the other hand, a person may be required to exit the bed at specified intervals since remaining in a lying position for a prolonged duration may result in bedsores or other forms of physical deterioration and undesirable health complications.

[0007] A person may be in a sleeping or a wakened state while undergoing monitoring of bed occupancy status. Information regarding bed occupancy, such as the frequency of waking up at night or leaving the bed to use the restroom, may provide an indication of sleep quality. An examination of bed occupancy states in conjunction with physiological measurements obtained over the course of a sleep session can suggest possible factors disturbing sleep and allow for mitigating the presence and effect of such disturbances in the future. Sleep testing or polysomnography equipment may be utilized for recording changes in physiological parameters during sleep. For example, a polysomnogram may measure brain activity eye movements, skeletal muscle activation, cardiac activity, oxygen saturation in the blood via pulse oximetry, as well as breathing function and respiratory effort. Such physiological measurements generally require cumbersome devices and sensors which need to be worn by or attached onto the body of the monitored person, and / or integrated on or near the bed. Continuous monitoring using such body attached devices is generally not feasible.

[0008] Various sensors may be employed to enable contactless monitoring. For example, a bed monitoring system may employ optical detection, such as using one or more visible-light cameras or infrared (IR) sensors. However, optical sensors typically 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. More generally, contactless sensors may be prone to errors, malfunctioning and gradual deterioration. Furthermore, individual sensors have a limited field of view and even the deployment of multiple sensors may not fully encompass an adequately broad coverage area. For example, one or more sensors may be improperly aligned or provide an undersized sensing zone, thereby covering only a limited portion of the bed and / or the body of the monitored person, and thus fail to preclude missed detections, such as of the person entering or exiting the bed.

[0009] An alternative approach is to devote a human operator to observe the subject while in the bed. This approach is labor intensive and time-consuming and is generally impractical for sustained periods.

[0010] Radar based systems for bed monitoring are known in the art. However, radar-based systems generally require a calibration process for each room, surface and / or subject prior to use. Such calibrations can be exceedingly cumbersome and arduous, particularly when monitoring multiple subjects in a given room, on different surfaces, or in changing settings that may not necessarily be known in advance.

[0011] SUMMARY

[0012] In accordance with one aspect of the present disclosure, there is thus provided a method for bed occupancy monitoring. The method includes receiving a millimeter-wave reflection radar signal from a monitoring area including a bed, and sampling the reflection radar signal and extracting a signal portion at a range of a subject of the bed, the signal portion consisting of an in-phase (I) signal component and a quadrature (Q) signal component. The method includes deriving a median zero-crossing over a selected time window from whichever one of the in-phase (I) signal component or the quadrature (Q) signal component has a larger magnitude, and determining a bed occupancy state of the bed based on the median zero-crossing. Determining the bed occupancy state includes comparing the median zero-crossing to a middle threshold, determining an empty bed state when the median zero-crossing is above the middle threshold, and determining an occupied bed state when the median zero-crossing is below the middle threshold. Determining the bed occupancy state may further include determining a transitional occupancy state by comparing the median zero-crossing to a low threshold and a high threshold, determining an entering bed state when the median zero-crossing is below the low threshold for at least a lag period, and determining an exiting bed state when the median zero-crossing is above the high threshold and at least one of: a waveform characteristic is detected in the in-phase signal component or the quadrature signal component; or the median zerocrossing exceeds the high threshold for at least a sustained period. The sustained period may be in the range of 5 to 20 seconds. The lag period may be in the range of 40 to 120 seconds. The selected time window may be in the range of 15 to 30 seconds. The method may further include issuing a notification relating to a determined bed occupancy state. The method may further include providing information relating to determined bed occupancy states over at least one time period. The method may further include pre-processing the extracted signal portion prior to deriving the median zero-crossing, where the pre-processing includes at least one of: bandpass filtering; normalization; and down-sampling. The radar signal may be at a frequency above 100 GHz. 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 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 method may further include predicting a future bed occupancy state using a machine learning model trained on historical bed occupancy data.

[0013] In accordance with another aspect of the present disclosure, there is thus provided a system for bed occupancy monitoring. The system includes a radar device and a processor. The radar device is configured to receive a millimeter-wave reflection radar signal from a monitoring area including a bed. The processor is configured to: sample the reflection radar signal and extract a signal portion at a range of a subject of the bed, the signal portion consisting of an in-phase (I) signal component and a quadrature (Q) signal component, derive a median zero-crossing over a selected time window from whichever one of the in-phase signal component or the quadrature signal component has a larger magnitude, and determine a bed occupancy state of the bed based on the median zero-crossing. Determining the bed occupancy state includes comparing the median zero-crossing to a middle threshold, determining an empty bed state when the median zero-crossing is above the middle threshold, and determining an occupied bed state when the median zero-crossing is below the middle threshold. Determining the bed occupancy state may further include determining a transitional occupancy state by comparing the median zero-crossing to a low threshold and a high threshold, determining an entering bed state when the median zero-crossing is below the low threshold for at least a lag period, and determining an exiting bed state when the median zero-crossing is above a high threshold and at least one of: a waveform characteristic is detected in the in-phase signal component or the quadrature signal component; or the median zerocrossing exceeds the high threshold for at least a sustained period. The sustained period may be in the range of 5 to 20 seconds. The lag period may be in the range of 40 to 120 seconds. The selected time window may be in the range of 15 to 30 seconds. The system may further include a notification unit, configured to issue a notification relating to a determined bed occupancy state. The system may further include a user interface, configured to provide information relating to determined bed occupancy states over at least one time period. The processor may be configured for pre-processing the extracted signal portion prior to deriving the median zero-crossing, wherein the pre-processing comprises at least one of: bandpass filtering; normalization; and down-sampling. The radar signal may be at a frequency above 100 GHz. 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 radar device may be 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 transm itted radar signal reflected from the body tissue of the subject. The processor may be configured to apply a machine learning model, to predict a future bed occupancy state based on historical bed occupancy data.

[0014] BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present disclosure will be understood and appreciated more fully from the following detailed description taken in conjunction with the drawings in which:

[0016] Figure 1 is a schematic illustration of a bed occupancy monitoring system, constructed and operative in accordance with an embodiment of the present disclosure;

[0017] Figure 2 is a schematic illustration of the radar device of the system of Figure 1 receiving reflections from a subject on a bed, constructed and operative in accordance with an embodiment of the present disclosure;

[0018] Figure 3 is an illustration of a state machine for determining a bed occupancy state, operative in accordance with an embodiment of the present disclosure;

[0019] Figure 4 is a graph of an exemplary reflection derived signal with l / Q signal components and depicting occurrences of a bed entry and a bed exit, operative in accordance with an embodiment of the present disclosure;

[0020] Figure 5A is a graph of an exemplary reflection derived signal with l / Q signal components and depicting a detection of a bed exit state, operative in accordance with an embodiment of the present disclosure;

[0021] Figure 5B is a graph of an exemplary median zero crossing waveform derived from the l / Q signals of Figure 5A, operative in accordance with an embodiment of the present disclosure;

[0022] Figure 6A is a graph of an exemplary reflection derived signal with l / Q signal components and depicting a detection of a bed entry state, operative in accordance with an embodiment of the present disclosure; Figure 6B is a graph of an exemplary median zero crossing waveform derived from the l / Q signals of Figure 6A, operative in accordance with an embodiment of the present disclosure;

[0023] Figure 7 is a flow diagram of a bed occupancy monitoring method, operative in accordance with an embodiment of the present disclosure;

[0024] Figure 8 is a flow diagram of a model training phase of a bed occupancy prediction method, operative in accordance with an embodiment of the present disclosure; and

[0025] Figure 9 is a flow diagram of a subject monitoring phase of a bed occupancy prediction method, operative in accordance with an embodiment of the present disclosure.

[0026] DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] The present disclosure may overcome the disadvantages of the prior art by providing a method and system for monitoring bed occupancy of a subject, with a high degree of accuracy and in a contact free manner, and without requiring a time-consuming calibration process for different environments, for different lying surfaces, and / or for different subjects.

[0028] 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.

[0029] 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. 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.

[0030] 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.

[0031] Certain features of the disclosure, 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 disclosure, 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.

[0032] 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. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1 , 2, 3, 4, 5, and 6. Whenever a numerical range is indicated herein, it is meant to include any cited numeral (fractional or integral) within the indicated range. For example, the phrases “ranging / ranges between” a first indicated number and a second indicated number and “ranging / ranges from” a first indicated number “to” a second indicated number are used herein interchangeably and are meant to include the first and second indicated numbers and all fractional and integral numerals there between.

[0033] 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.

[0034] 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.

[0035] The term “subject” is used herein to refer to an individual upon which the method or system of the present disclosure is performed, such as a person undergoing a monitoring of bed occupancy status. 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.

[0036] The terms “user” and “operator” are used interchangeably herein to refer to any individual person or group of persons using or operating at least one aspect of a method or system of the present disclosure.

[0037] The term “bed” is used herein generally to refer to any type of surface or platform upon which one or more subjects may be positioned in a recumbent or lying posture (i.e., substantially horizontal to the ground) when supported by the bed, as well as near lying postures, including surfaces or platforms that can be tilted or reclined to allow for subject positioning at different angles, such as a reclinable seat that can allow for sitting, lying, or reclining postures. Examples of a bed may include but are not limited to: a bed; a couch; a sofa; a cot; a crib; a divan; a mattress; a box-spring; and the like. A bed may include multiple portions, such as a first portion configured to support a back of a subject (e.g., a backrest) and a second portion configured to support a head of a subject (e.g., a headrest).

[0038] Reference is now made to Figure 1 , which is a schematic illustration of a bed occupancy monitoring 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 user interface 116, a notification unit 117, and a database 118. Processor 114 is communicatively coupled with radar device 112, with user interface 116, with notification unit 117, and with database 118.

[0039] Radar device 112 is configured to transmit a radar signal 122 to a monitoring area, such as to at least a portion of a bed 130 where an intended subject 120 may be occupying. 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 a subject 120 present in the monitoring area. 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 bed 130, constructed and operative in accordance with an embodiment of the present disclosure. System 110 may be applied for monitoring a bed occupancy of a bed 130 configured for supporting a subject 120, such as in a lying or recumbent posture with the subject body aligned substantially horizontally or at an incline, such as to enable sleeping or resting by subject 120 on bed 130. Radar device 112 may be positioned in the vicinity of bed 130, such as mounted or coupled to a structure above bed 130 (e.g., a wall, a ceiling, a post) and directed to a supporting surface of bed 130, such that a field of view of radar device 112 encompasses at least a portion of bed 130 where subject 120 may be present. In one example, the field of view of radar device 112 encompasses substantially an entirety of bed 130. In another example, a plurality of radar devices 112 may be deployed, such that each individual radar device 112 is directed to a respective field of view (FOV) in the vicinity of bed 130, where the FOVs of multiple radar devices 112 may encompass an entirety of bed 130. Radar device 112 may be positioned at a predetermined distance from bed 130, such as within a few meters.

[0040] 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 of the present disclosure, the transmitted and reflected radar signals 122, 124 is in the Terahertz (THz) frequency band, where the term “Terahertz (THz)” as used herein encompasses Terahertz and sub-Terahertz radiation corresponding to sub-millimeter 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).

[0041] 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 transmit 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 bed 130 on which subject 120 may be positioned.

[0042] 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 determine a bed occupancy status, as will be elaborated upon further hereinbelow.

[0043] User interface 116 allows a user to receive information and to control parameters or settings associated with system 110. For example, user interface 116 may include a display screen configured to present visual content. User interface 116 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 116 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.

[0044] Notification unit 117 is configured to provide alerts or notifications relating to the operation of system 110. For example, notification unit 117 may issue a first alert when a bed entry state is detected, and a second alert when a bed exit state is detected, as well as additional indications relating to other occupancy statuses of a monitored bed. Notification unit 117 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 117 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 117 may optionally be integrated with user interface 116 or may be a separate unit. Notification unit 117 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 web-based communication network; a supervisor or central monitoring station (e.g., associated with a health care facility), and the like. A plurality of notification units 117 may be employed for providing alerts at different locations.

[0045] 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.

[0046] 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. 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.

[0047] The operation of system 110 will now be described in general terms, followed by specific examples. Radar device 112 transmits a coherent radar signal 122, such as a frequency-modulated continuous wave (FMCW) radar signal in the THz frequency band, to a monitored bed 130. Radar device 112 may receive a corresponding reflected radar signal 124 from monitored bed 130, such as from a body part of a subject 120 on bed 130, where reflection radar 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 incident on different body parts of subject 120, such as an anterior body area (e.g., the chest), or a posterior body area (e.g., the back). 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, and measures a phase difference between the transmitted signal 122 and received signal 124 to determine the distance or range traversed by the propagating radar signal, so as to extract the 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 corresponding to the subject range. For example, if operating in FMCW mode the reflected radar signal is sampled at a selected rate, e.g., 500 Hz, such that 500 times per second a vector of multiple samples (e.g., 128 samples) is collected (e.g., providing 64kHz samples per second). This received signal of multiple samples may undergo a Fourier transform, such that each transformed sample is respective of a range. Of these transformed samples, an individual sample corresponding to the subject range is extracted, resulting in a collection of values in accordance with the sampling rate (e.g., 500 extracted subject range samples per second). A determination of the subject range 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”.

[0048] The extracted signal portion (e.g., the output of the FFT, at the subject range) collected over predefined time intervals (e.g., 500 Hz) and consisting of an in-phase (I) component and a quadrature (Q) component, may optionally undergo pre-processing. For example, a nonlinear filtering or mapping operation may be applied, such as to produce a complex-valued (ballistocardiograph) signal providing a time-domain representation of ballistic forces associated with cardiac activity and pulmonary activity. The pre-processing may include bandpass filtering to remove very low frequencies and very high frequencies (e.g., ranging between 0.05Hz and 3.33Hz), such that frequencies pertaining to vital signs of the subject, such as respiration and heartrate and various harmonics, derivatives, traces and effects of physiological phenomena on the body, remain in the signal. Other pre- processing operations may include normalization (such as by applying a gain control function), and down-sampling to a selected sampling rate (e.g., ranging from 500Hz to 10Hz). The pre-processing operations may be interchangeable and implemented in different orders.

[0049] The reflection radar derived signal with l / Q waveforms is then processed to extract a median zero-crossing (MZC) measure over a selected time window, such as a duration of between 15 and 30 seconds. The MZC signal waveform may be calculated from the maximum of the I and Q signals. The extracted MZC is then compared with established threshold values to determine a bed occupancy status. In particular, the MZC may be compared with a “high threshold”, a “middle threshold” and a “low threshold”. An initial base assessment of bed occupancy may be obtained by applying a first binary classification. Specifically, if the MZC exceeds the middle threshold, then the monitored bed may be determined to be empty (i.e., an “empty bed state” or “bed empty state”), whereas if the MZC is below the middle threshold, then the monitored bed may be determined to be occupied (i.e., an “occupied bed state” or “bed occupied state”). Additional conditions may be evaluated to indicate transitions between the aforementioned base occupancy states. A transition from a bed occupied state to a bed empty state, reflecting a subject exiting the bed (i.e., an “exiting bed state” or “bed exit state”), may be triggered when the MZC exceeds the high threshold along with a detection of a particular waveform characteristic of the l / Q signals, or alternatively when the MZC exceeds the high threshold for a sustained duration (e.g., for at least 5-20 seconds). Conversely, a transition from a bed empty state to a bed occupied state, reflecting a subject entering the bed (i.e., an “entering bed state” or “bed entry state”), may be initiated when the MZC drops beneath the low threshold for a selected minimum interval (e.g., for at least 40-120 seconds) referred to herein as a “lag period”. The thresholds may be determined empirically, such as based on testing results.

[0050] Reference is made to Figure 3, which is an illustration of a state machine 150 for determining a bed occupancy state, operative in accordance with an embodiment of the present disclosure. Processor 114 may determine a bed occupancy state of a monitored bed 130 using state machine 150, based on processing of l / Q component signals derived from reflection radar signal 124. State machine 150 may include four bed occupancy states: an empty bed state 152 (S1 ); an entering bed state 154 (S2); an occupied bed state 156 (S3); and an exiting bed state 158 (S4). Empty and occupied states (S1 , S3) may be considered “base states”, whereas entering and exit states (S2, S4) may be considered “transitional states”. If no reflection radar signal 124 is received, or when an extracted MZC of the l / Q signals is above a middle threshold value, then a base state of empty (S1 ) may be established. Following empty bed state 152 (S1 ), a next possible state may be an entering bed state 154 (S2) in which a subject 120 is entering bed 130. Detection of the extracted MZC dropping below a low threshold for a minimum duration may indicate that an entering bed state 154 has occurred, causing state machine 150 to transition from state S1 to state S2. Following bed entry state 154 (S2), a next possible state may be an occupied bed state 156 (S3) in which subject 120 is occupying bed 130. Specifically, monitored bed 130 enters a base state of occupied (S3) immediately following the transitional state of bed entry (S2). When an extracted MZC of the l / Q signals is below the middle threshold value, then an occupied bed state 156 (S3) may be established, causing state machine 150 to transition from state S2 to state S3. Following bed occupied state 156 (S3), a next possible state may be an exiting bed state 158 (S4) in which subject 120 is exiting bed 130. Detection of an extracted MZC exceeding a high threshold along with a particular waveform characteristic of the l / Q signals, or alternatively a detection of the extracted MZC exceeding the high threshold for a sustained duration, may indicate that exiting bed state 158 has been entered, causing state machine 150 to transition from state S3 to state S4. Following bed exit state 158 (S4), a next possible state may be an empty bed state 152 (S1 ), which occurs after the subject 120 has left bed 130. Specifically, monitored bed 130 enters a base state of empty (S1 ) immediately following the transitional state of bed exit (S1 ).

[0051] Reference is made to Figure 4, which shows a graph, generally referenced 160, of an exemplary reflection derived signal with l / Q signal components and depicting occurrences of a bed entry and a bed exit, operative in accordance with an embodiment of the present disclosure. Graph 160 illustrates a radar reflection derived signal with an in-phase waveform 162 and a quadrature waveform 163, plotted as a function of time. At a first time instant (ti), corresponding to an elapsed time of approximately 300 seconds (sec), the characteristics of waveforms 162, 163 change substantially (e.g., are characterized by rapid fluctuations), which may be indicative of a bed entry state. At a second time instant (t2), corresponding to an elapsed time of approximately 600 sec, the characteristics of waveforms 162, 163 again change substantially (e.g., are characterized by a relatively constant value, similar to an initial pattern appearing prior to time instant ti ), which may be indicative of a bed exit state.

[0052] Reference is made to Figures 5A and 5B. Figure 5A shows a graph, generally referenced 170, of an exemplary reflection derived signal with l / Q signal components and depicting a detection of a bed exit state, operative in accordance with an embodiment of the present disclosure. Graph 170 illustrates a radar reflection derived signal with an in-phase waveform 172 and a quadrature waveform 173, plotted as a function of time. Figure 5B shows a graph, generally referenced 180, of an exemplary median zero crossing waveform, referenced 181 , derived from the l / Q signals of Figure 5A, operative in accordance with an embodiment of the present disclosure. Graph 180 further depicts a middle threshold, referenced 186, and having an exemplary value of 0.3, and a high threshold, referenced 187, and having an exemplary value of 0.4. At a first time instant (ti) during an elapsed time between 600 to 700 sec, median zero crossing (MZC) waveform 181 exceeds the value of middle threshold 186. MZC waveform 181 continues to increase and subsequently exceeds the value of high threshold 187. At a second time instant (t2), quadrature waveform 173 exhibits a predetermined waveform characteristic or pattern, referenced 175, such as a relatively gradual increase followed by a gradual decrease and then another increase (e.g., somewhat resembling a sinusoidal wave pattern). It is noted that the predetermined waveform characteristic may appear on either of the in-phase or quadrature signals (e.g., may be calculated on the maximum of the I and Q signals), and is depicted on graph 180 for exemplary purposes on quadrature waveform 173. Accordingly, a bed exit state may be determined to have occurred at second time instant t2. If predetermined waveform characteristic 175 is not identified in the l / Q signals 172, 173, then a bed exit state may be established when MZC waveform 181 has exceeded high threshold 187 for a sustained duration, such as at a third time instant (ts). Reference is made to Figures 6A and 6B. Figure 6A shows a graph, generally referenced 190, of an exemplary reflection derived signal with l / Q signal components and depicting a detection of a bed entry state, operative in accordance with an embodiment of the present disclosure. Graph 190 illustrates a radar reflection derived signal with an in-phase waveform 192 and a quadrature waveform 193, plotted as a function of time. Figure 6B shows a graph, generally referenced 200, of an exemplary median zero crossing waveform, referenced 201 , derived from the l / Q signals of Figure 6A, operative in accordance with an embodiment of the present disclosure. Graph 200 further depicts a low threshold, referenced 205, and having an exemplary value of 0.2, and a middle threshold, referenced 206, and having an exemplary value of 0.3. At a first time instant (ti) during an elapsed time between 300 to 400 sec, median zero crossing (MZC) waveform 201 drops below the value of middle threshold 206. At a second time instant (t2) subsequent to the first time instant ti, median zero crossing (MZC) waveform 201 drops below the value of low threshold 205. At a third time instant (ts) following a lag period 208 of predetermined minimum duration (e.g., between 40-120 seconds) after the second time instant t2, MZC waveform 208 has remained beneath low threshold 205 for a sustained duration, such that a bed entry state is determined to have occurred at third time instant ts.

[0053] Reference is now made to Figure 7, which is a flow diagram of a bed occupancy monitoring method, operative in accordance with an embodiment of the present disclosure. In a step 222, a reflection radar signal is received from a monitoring area with a bed. Referring to Figures 1 and 2, radar device 112 transmits a coherent radar signal 122, such as a FMCW radar signal in the THz frequency band, to the vicinity of a monitored bed 130. Radar device 112 receives a corresponding reflected radar signal 124, such as from a body part of a subject 120 on bed 130, such that reflection radar signal 124 contains information relating to micro-displacements in the skin associated with the cardiac cycle and pulmonary activity of subject 120. It is noted that the monitored bed 130 may be occupied by a plurality of subjects 120.

[0054] In step 224, the reflected signal is sampled and a signal portion at the range of a bed 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, and measures a phase difference between the transmitted signal 122 and received signal 124 to determine the distance traversed by the radar signal, so as to extract the 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 corresponding to the subject range. For example, if operating in FMCW mode the reflected radar signal is sampled at a selected rate, e.g., 500 Hz, such that 500 times per second a vector of multiple samples (e.g., 128 samples) is collected (e.g., providing 64kHz samples per second). This received signal (e.g., of 128 samples) undergoes a Fourier transform, such that each transformed sample is respective of a range. Of these transformed samples, an individual sample corresponding to the subject range is extracted, resulting in a collection of values in accordance with the sampling rate (e.g., 500 extracted subject range samples per second). Pre-processing operations may optionally be applied to the resultant signal (e.g., nonlinear filtering, bandpass filtering, normalization, downsampling, and the like), such as to enhance or facilitate subsequent signal processing.

[0055] In step 226, a median zero crossing is derived over a selected time window of an l / Q signal component. Referring to Figures 1 and 2, processor 114 processes the in-phase and quadrature signals associated with reflection radar signal 124 to extract at least one median zero-crossing (MZC) measure over a respective duration (e.g., 15-30 seconds). The MZC may be derived from the maximum (larger) one of the l / Q signal components. For example, referring to Figures 5A and 5B, an MZC waveform 181 is derived from in-phase signal 172 or quadrature signal 173 of graph 170, such as based on the larger of signals 172, 173.

[0056] In step 228, a bed occupancy state of the bed is determined based on the median zero crossing. Referring to Figures 1 and 2, processor 114 determines a bed occupancy status of monitored bed 130 based on analysis of the in-phase and quadrature signal components and examining the extracted MZC waveform in relation to established threshold values. Step 228 may include sub-procedures 230, 232. In sub-procedure 230, a base occupancy state is determined, where an empty bed state is determined when the median zero crossing is above a middle threshold, and where an occupied bed state is determined when the median zero crossing is below the middle threshold. Referring to Figures 1 , 2 and 3, processor 114 determines that a monitored bed 130 is in an empty bed state 152 when an MZC derived from l / Q signals associated with reflection radar signal 124 is above a middle threshold value, and determines that bed 130 is in an occupied bed state 156 when the MZC is below the middle threshold value. For example, referring to Figure 5B, when MZC waveform 181 is above middle threshold 186 (e.g., at an elapsed time after 700 sec) then an empty bed state 152 is established, and when MZC waveform 181 is below middle threshold 186 (e.g., at an elapsed time before 600 sec) then an occupied bed state 156 is established. Monitored bed 130 may enter an empty bed state (S1 ) after the occurrence of a transitional bed exit state (S4), and monitored bed 130 may enter an occupied bed state (S3) after the occurrence of a transitional bed entry state (S2).

[0057] In sub-procedure 232, a transitional occupancy state is determined, where an entering bed state is determined when the median zero crossing is below a low threshold for at least a minimum (lag) period, and where an exiting bed state is determined when the median zero crossing is above a high threshold and at least one of: a waveform characteristic is detected, or the high threshold is exceeded for a sustained period. Referring to Figures 1 , 2 and 3, processor 114 determines that monitored bed 130 is in an entering bed state 154 when an MZC derived from l / Q signals associated with reflection radar signal 124 drops below a low threshold value for at least a predetermined duration (e.g., between 40-120 seconds). For example referring to Figure 6B, after MZC waveform 201 falls below low threshold 205 (at time t2) and has remained beneath low threshold 205 for at least a lag period 208 (e.g., at time ts), then an entering bed state 154 is established. For another example, referring to Figure 5B, after MZC waveform 181 exceeds high threshold 187 (e.g., at an elapsed time after 600 sec), and when a predetermined waveform characteristic 175 of quadrature waveform 173 is detected (e.g., at time t2), then an exiting bed state 158 is established. Monitored bed 130 may enter a transitional bed entry state (S2) after the occurrence of an empty bed state (S1 ), and monitored bed 130 may enter a transitional bed exit state (S4) after the occurrence of an occupied bed state (S3).

[0058] Sub-procedure 230, 232 may be performed iteratively, where sub-procedure 232 may be performed before and / or after sub-procedure 230, such as in accordance with state machine 150 (Fig.3).

[0059] In step 234, a notification relating to a determined bed occupancy state is issued. Referring to Figures 1 , 2 and 3, notification unit 117 issues a notification or alert relating to a determined bed occupancy state of monitored bed 130. The notification may be provided via a visual indication and / or an audible indication. For example, notification unit 117 may issue an alert when the bed occupancy status transitions from a bed occupied state (S3) to a bed empty state (S1 ), such as during a bed exit state (S4). An issued notification may be linked to one or more conditions, which may be configured or modified by a user. For example, an alert may be issued if a subject 120 remains out of bed 130 for a predefined minimum time interval (i.e. , if an empty bed state S1 occurs for at least the predefined time interval). For another example, an alert may be issued if a subject 120 leaves bed 130 more often than a predefined maximum limit during a certain time period (i.e., if the number of exiting bed states 158 occurring during the time period exceeds the predefined limit). 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 a bed exit of a subject 120 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 a bed entry of a subject 120 with minimal health issues) may be issued as a low volume or low frequency audible sound and / or a graphical symbol with low intensity.

[0060] More generally, system 110 may provide a report, such as via user interface 116, relating to a bed occupancy status of monitored bed 130. The report may include a visual representation of determined bed occupancy states, such as over a selected monitoring period, and optional supplementary information about subject 120 and bed 130. The report may include characteristics of a determined bed occupancy state, such as: the duration of the occupancy state, the frequency of occurrences of each occupancy state during a given time interval, a confidence level of the occupancy state (e.g., reflecting a reliability of the determination), and types of alerts provided during the time interval. The report may further include various statistics associated with the determined bed occupancy states, such as historical data obtained at previous dates and times, or bed occupancy states determined for similar types of beds or subjects. The information and statistics may be presented over a selected duration, such as depicting dynamic changes over time. For example, a report may be compiled presenting a total number of bed exit states each night by a subject 120 over a given time period (e.g., several days, weeks, or months), which may provide insight into their behaviour and wellbeing. Furthermore, notification unit 117 may issue a notification or alert based on deviations of bed occupancy states relative to statistical or historical data, which may provide early warning signs of changes in a condition of subject 120.

[0061] The method of Figure 7 is generally implemented in an iterative manner, such that at least some of the steps and sub-procedures are performed repeatedly, in order to provide for a dynamic determination of bed occupancy status of a monitored bed in real-time.

[0062] According to an aspect of the present disclosure, a prediction of a bed occupancy status may be established using machine learning processes. Reference is made to Figure 8, which is a flow diagram of a model training phase 250 of a bed occupancy prediction method, operative in accordance with an embodiment of the present disclosure. A reflection radar signal 252 is obtained using radar device 112 from a reference bed 230 where a subject 220 may be present. Reflection radar signal 252 is sampled and filtered to extract a signal portion 254 at the subject range having in-phase (I) and quadrature (Q) components. The extracted signal 254 with l / Q components, following optional pre-processing, is divided into discrete segments 258 of a selected temporal duration. A detected bed occupancy state 262 is obtained from reference measurement bed 230 using one or more external detection means, such as using a motion or proximity sensor and / or a human operator. A training sample is formed by assigning each signal segment a respective label based on a detected bed occupancy status 262 during the time period of that signal segment 258. This process is repeated for a large number of reference measurements of one or more beds 230 to generate a training dataset 266 made up of a collection of training samples, each training sample made up of a signal segment 258 and assigned label (i.e. , detected bed occupancy state 262). A group of training samples may be obtained for a defined time window leading up to a respective bed occupancy state (e.g., immediately prior to a bed entry state S2 and immediately prior to a bed exit state S4). The collection of training samples may include samples obtained from different sources, such as different types of reference measurement beds employing different occupancy detection modalities, which may provide variation to augment the training process. The collection of training samples 266 is fed into a training model 268 that utilizes machine learning techniques to produce a bed occupancy prediction model 270 that can be applied on new input datasets. The generated bed occupancy prediction model 270 may be dynamically updated and subsequently applied to predict a bed occupancy state (and supplementary statistics) of a monitored bed during a subsequent monitoring phase.

[0063] Training model 268 may analyze the dataset of training samples 266 using at least one machine learning process, to implicitly identify different patterns and create models for predicting bed occupancy states. The machine learning process may apply machine learning techniques to analyze the training data in order to produce mapping functions that can be used by prediction model 270 for classifying additional instances of new datasets 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. A generated prediction model 270 may be iteratively updated and improved based on new information, such as accounting for subsequent successful or unsuccessful bed occupancy predictions, and additional collected training data. Simulations of numerous collected data and bed occupancy predictions may be applied to enhance the reliability and accuracy of the prediction model 270. Updated models may provide optimal formulas and weighting metrics for different variables or classification features. As more information and statistics are accumulated, prediction model 270 may be further refined to improve predictive capabilities.

[0064] Reference is made to Figure 9, which is a flow diagram of a subject monitoring phase 350 of a bed occupancy prediction method, operative in accordance with an embodiment of the present disclosure. A monitoring sample of a monitored bed 330 may be obtained according to a similar sequence described hereinabove in the context of model training phase 250. In particular, a reflection radar signal 352 is obtained 330 using radar device 112 from a monitored bed 330 where a subject 320 may be present. Reflection radar signal 352 is sampled and filtered to extract a signal portion 354 at the subject range having in-phase (I) and quadrature (Q) components. The extracted signal 354 with l / Q components, following optional pre-processing, is divided into discrete segments of a selected temporal duration. The segments, accumulated over the selected sample duration, make up monitoring samples 358 which are fed into bed occupancy prediction model 270. Monitoring samples 358 may be provided to bed occupancy prediction model 270 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 270. Alternatively, samples 358 may be delivered in parallel to bed occupancy prediction model 270 (“offline”) after waiting to acquire and segment a longer input signal. The particular segments of monitoring samples 358 that are processed may be dynamically selected. It is noted that consecutive segments may overlap, whereby consecutive segments of a certain duration may overlap by a certain amount.

[0065] Bed occupancy prediction model 270 applies machine learning processes to the monitoring samples 358 to predict a bed occupancy state in a defined time window, such as based on correlating data patterns identified in reference signal segments of training samples 266 during earlier training of model 270. For example, a bed exit state (S4) may be predicted based on patterns identified (by training model 268) in reference signal segments of training samples 266 obtained during a time window leading up to a known (detected) bed exit state during the training process 250. Similarly, bed occupancy prediction model 270 may correlate certain data patterns with the occurrence of other bed occupancy states for predicting future states based on similar patterns observed within the defined time window leading up to that state. A notification or alert may be issued in accordance with a predicted bed occupancy state.

[0066] It is appreciated that the disclosed embodiments may provide a reliable and accurate assessment of bed occupancy status of a monitored area over time. Further statistics obtained over multiple sessions and associated historical data can provide a more comprehensive evaluation and form the basis for providing targeted recommendations for treating subjects. The disclosed embodiments may not require components to be in direct physical contact with the subject of a monitored bed, such that there is no need for coupling a sensor or other device to the body of a subject before a monitoring session, thus saving time and minimizing discomfort. Furthermore, a subject on a monitored bed may 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 utilized to enable measuring and identifying multiple subjects concurrently. The disclosed system does not require costly equipment and has relatively few components and is relatively straightforward to operate and maintain. The bed occupancy status determination and / or prediction of the disclosed method and system may be implemented in a wide variety of locations (e.g., rooms or facilities), and beds (e.g., different types and sizes of beds), and different types of subjects (e.g., regardless of age, height, weight, or other physical characteristics), without requiring a time-consuming and cumbersome calibration process prior to each particular monitoring session depending on the type of location, type of bed, or type of subject. A machine learning based model may be applied to provide reliable and accurate bed occupancy status predictions, which can be iteratively refined to improve predictive capabilities based on new information.

[0067] 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 and improve treatment in an eldercare facility.

[0068] While certain embodiments of the disclosed subject matter have been described, so as to enable one of skill in the art to practice the disclosed subject matter, 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 bed occupancy monitoring, the method comprising: receiving a millimeter-wave reflection radar signal from a monitoring area comprising a bed; sampling the reflection radar signal and extracting a signal portion at a range of a subject of the bed, the signal portion consisting of an in-phase (I) signal component and a quadrature (Q) signal component; deriving a median zero-crossing over a selected time window from whichever one of the in-phase signal component or the quadrature signal component has a larger magnitude; determining a bed occupancy state of the bed based on the median zero-crossing, wherein determining the bed occupancy state comprises: comparing the median zero-crossing to a middle threshold; determining an empty bed state when the median zero-crossing is above the middle threshold; and determining an occupied bed state when the median zero-crossing is below the middle threshold.

2. The method of claim 1 , wherein determining the bed occupancy state further comprises determining a transitional occupancy state by: comparing the median zero-crossing to a low threshold and a high threshold; determining an entering bed state when the median zero-crossing is below the low threshold for at least a lag period; and determining an exiting bed state when the median zero-crossing is above the high threshold and at least one of: a waveform characteristic is detected in the in-phase signal component or the quadrature signal component; or the median zero-crossing exceeds the high threshold for at least a sustained period.

3. The method of claim 2, wherein the sustained period is in the range of 5 to 20 seconds.

4. The method of claim 2, wherein the lag period is in the range of 40 to 120 seconds.

5. The method of claim 1 , wherein the selected time window is in the range of 15 to 30 seconds.

6. The method of claim 1 , further comprising issuing a notification relating to a determined bed occupancy state.

7. The method of claim 1 , further comprising providing information relating to determined bed occupancy states over at least one time period.

8. The method of claim 1 , further comprising pre-processing the extracted signal portion prior to deriving the median zero-crossing, wherein the pre-processing comprises at least one of: bandpass filtering; normalization; and down-sampling.

9. The method of claim 1 , wherein the radar signal is at a frequency above 100 GHz.

10. The method of claim 1 , wherein the radar signal is a frequency-modulated continuous-wave (FMCW) radar signal.11 . The method of claim 1 , wherein the signal portion is sampled at a sampling rate of 500 Hz.

12. 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.

13. The method of claim 1 , further comprising predicting a future bed occupancy state using a machine learning model trained on historical bed occupancy data.

14. A system for bed occupancy monitoring, the system comprising: a radar device, configured to receive a millimeter-wave reflection radar signal from a monitoring area comprising a bed; and a processor, configured to: sample the reflection radar signal and extract a signal portion at a range of a subject of the bed, the signal portion consisting of an in-phase (I) signal component and a quadrature (Q) signal component; derive a median zero-crossing over a selected time window from whichever one of the in-phase signal component or the quadrature signal component has a larger amplitude; determine a bed occupancy state of the bed based on the median zero-crossing, wherein determining the bed occupancy state comprises: comparing the median zero-crossing to a middle threshold; determining an empty bed state when the median zero-crossing is above the middle threshold; and determining an occupied bed state when the median zerocrossing is below the middle threshold.

15. The system of claim 14, wherein determining the bed occupancy state further comprises determining a transitional occupancy state by: comparing the median zero-crossing to a low threshold and a high threshold; determining an entering bed state when the median zero-crossing is below the low threshold for at least a lag period; and determining an exiting bed state when the median zero-crossing is above the high threshold and at least one of: a waveform characteristic is detected in the in-phase signal component or the quadrature signalcomponent; or the median zero-crossing exceeds the high threshold for at least a sustained period.

16. The system of claim 15, wherein the sustained period is in the range of 5 to 20 seconds.

17. The system of claim 15, wherein the lag period is in the range of 40 to 120 seconds.

18. The system of claim 14, wherein the selected time window is in the range of 15 to 30 seconds.

19. The system of claim 14, further comprising a notification unit, configured to issue a notification relating to a determined bed occupancy state.

20. The system of claim 14, further comprising a user interface, configured to provide information relating to determined bed occupancy states over at least one time period.

21. The system of claim 14, wherein the processor is configured for pre-processing the extracted signal portion prior to deriving the median zero-crossing, wherein the pre-processing comprises at least one of: bandpass filtering; normalization; and down-sampling.

22. The system of claim 14, wherein the radar signal is at a frequency above 100 GHz.

23. The system of claim 14, wherein the radar signal is a frequency-modulated continuous-wave (FMCW) radar signal.

24. The system of claim 14, wherein the signal portion is sampled at a sampling rate of 500 Hz.

25. The system of claim 14, 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.

26. The system of claim 14, wherein the processor is configured to apply a machine learning model trained on historical bed occupancy data to predict a future bed occupancy state.-SO-

Images