Inductive sensing system and method

Abstract

A system (8) and method for extracting component signals relating to different physiological phenomena in a body from a sensed induction signal. A resonator circuit (10) is oscillated at a frequency to generate an alternating electromagnetic field which is applied to a body under investigation. This field induces a secondary eddy current in the body which interacts with the primary magnetic field and at least changes the frequency and amplitude of the resonator circuit oscillating current. These changes in the current characteristics, in particular the frequency and amplitude, are measured and provide first and second input signals. Embodiments of the invention provide a system (8) or method arranged to receive these input signals. A plurality of different composite or fused signals are then generated by the system, each signal being formed from a different linear combination ratio of the two input signals. These signals are then evaluated using a signal selection procedure to identify the best candidate signal or signals for providing a measure or indication of a particular physiological phenomenon. This can be based on, for example, predefined selection criteria relating to signal characteristics of the candidate signals.

Description

Technical Field

[0001] The present invention relates to an inductive sensing system, particularly for detecting and separating signal components indicating different physiological phenomena. Background Technology

[0002] There is often a need to measure the mechanical movement and dynamic changes of internal body structures such as the heart, lungs, or arteries. For example, it is useful to measure the periodic changes in the internal volume or size of the chambers of the heart or lungs or the arteries (e.g., the arterial volume changes with the cardiac cycle).

[0003] Sensors that measure mechanical activity are sometimes called wave-recording sensors (wave-recording biometric sensors in the clinical field). Some examples of wave-recording biometric sensors include accelerometer-based biosensors, transthoracic impedance biosensors, radar-based biosensors, capacitance sensors, and photoplethysmography (PPG) sensors.

[0004] Magnetic inductive sensors also have the potential to be used as bioassay sensors for sensing mechanical activity. The working principle of inductive sensing is based on Faraday's law. An oscillating primary magnetic field is generated by a generating loop antenna, and this induces eddy currents in the tissue radiated by the signal via Faraday's law. These eddy currents generate a secondary magnetic field. The total magnetic field is then the superposition of the primary and secondary magnetic fields. Changes induced in the electrical characteristics (antenna current) of the generating antenna can be measured, and these are used to infer the secondary magnetic field and thus the properties of the stimulated tissue.

[0005] Inductive sensing offers the potential for simple, contactless measurement of the mechanical activity of the heart and lungs, or the mechanical activity of blood vessels such as the radial artery in the human arm.

[0006] A significant drawback of biosensors (including inductive sensors and other types of sensors) is that it is currently very difficult to distinguish sensing signals originating from different physiological sources.

[0007] For example, it is very difficult to distinguish signal elements related to the mechanical activity of the heart (e.g., heart pulse) from those related to the mechanical activity of the lungs (e.g., breathing) within a single composite return signal.

[0008] In known systems, this is often done by assuming the heart rate is greater than the respiratory rate. In this way, different signal components can be distinguished and separated based on frequency.

[0009] However, clinical ranges for heart rate and pulse rate may overlap. For example, the frequency of exceptionally high respiratory rates may overlap with the frequency of exceptionally low pulse rates, and vice versa. Therefore, a sensed (high) respiratory rate may be incorrectly interpreted as a low heart rate by existing technology systems, and vice versa, leading to incorrect clinical diagnoses and interventions.

[0010] For example, in neonatal monitoring, apnea is often not detected by impedance-based measurements because the cardiac contractile characteristics of the condition are incorrectly interpreted as the patient’s respiratory rate.

[0011] Another related problem with known systems is that motion artifacts are difficult to distinguish from genuine biophysical signals, such as pulse or respiration. Distinguishing motion artifacts from genuine signals is often done by assuming that the frequency of the artifact differs from the frequency of the biometric signal and / or that the waveform of the artifact differs from the waveform of the biometric signal (e.g., in terms of shape characteristics). However, these methods are often unsatisfactory because both the frequency and waveform of the signal artifact often resemble the frequency and waveform of the biometric signal.

[0012] For example, if a patient's walking cadence is close to their pulse rate (which is usually the case), then the pulse rate measurement becomes unreliable. Furthermore, even when the frequencies differ, a patient's walking cadence may still be interpreted as pulse rate, for example, due to the very similar waveforms of the two. This leads to incorrect heart rate measurements.

[0013] Therefore, there is a need for an improved method of inductive sensing that can more reliably distinguish different physiological signals from each other and from signal artifacts.

[0014] WO2018 / 127488A1 discloses a magnetic induction sensing device including a loop antenna for inductively coupling with an electromagnetic (EM) signal emitted from a medium in response to stimulation of the medium by an electromagnetic excitation signal. In one embodiment, both antennas are connected to associated oscillators for driving the antennas at different corresponding frequencies f1 and f2. The electromagnetic signals received back at each antenna in response to the generated electromagnetic excitation signal will also be at the corresponding frequencies f1 and f2, and these electromagnetic signals are mixed with each other by a mixer and an applied low-pass filter configured to allow the differential frequencies to pass through. The signal is then passed to another signal processing element, such as a counter.

[0015] WO2018 / 127482A1 discloses a magnetic induction sensing system for sensing electromagnetic signals emitted from a body in response to an applied electromagnetic excitation signal. The electromagnetic signal is generated and sensed by the same ring resonator, which includes a single-turn ring antenna and a tuning capacitor. The ring antenna and a signal generation unit for exciting the resonator to generate the excitation signal are configured such that the ratio between the radial frequency of the generated electromagnetic excitation signal and a reference frequency of the antenna is optimized, where the reference frequency is a frequency of the generated excitation signal (wave) whose wavelength matches the circumference of the antenna. The ratio corresponding to the regularized radial frequency of the generated excitation signal is maintained between 0.025 and 0.50. Summary of the Invention

[0016] According to an example based on aspects of the invention, a system is provided for processing electromagnetic signals returned from a body in response to an electromagnetic excitation signal applied to the body in an inductive sensing process. The system is configured to receive a signal input indicating the sensed return signal, the return signal corresponding to a signal sensed at the loop antenna of the resonator circuit based on a change in the electrical characteristics of the resonator circuit detected when the circuit is driven to generate the excitation signal; The system is configured to implement a signal extraction procedure, wherein the system is configured to: A first input signal is detected from the sensed return signal, the first input signal being based on the frequency of the sensed return signal. A second input signal is detected from the sensed return signal, the second input signal being based on the sensing amplitude of the sensed return signal. The application signal generation program includes generating multiple candidate signals, each candidate signal being formed by different linear combinations of the first input signal and the second input signal, and A signal selection procedure is applied to select one of the candidate signals, the selection procedure being based on predefined criteria related to one or more signal characteristics of the input signal, the criteria being configured to isolate signals related to specific physiological sources in the body, the selected signal forming an output signal.

[0017] The system can be configured, for example, to communicate with an inductive sensing device during use, the inductive sensing device including a resonator circuit including a loop antenna arranged to be driven to generate an electromagnetic excitation signal, and including a signal sensing unit for sensing the return signal from the body based on detecting a change in the electrical characteristics of the resonator circuit.

[0018] The system may include, for example, a processing unit or controller for receiving the signal input indicating the sensed return signal and for executing the signal extraction procedure.

[0019] Embodiments of the present invention provide a method for extracting signal components from different physiological sources from a measured inductive signal. This is based on collecting at least two measured input signals, the first based on frequency variations of a resonator circuit and the second based on amplitude variations. The relative amounts of a given physiological signal present in each of the first and second input signals will typically be different. Therefore, by combining (or fusing) the two input signals at specific different ratios, a resulting output signal can be obtained, wherein unwanted physiological signal components are suppressed and desired signal components are emphasized or enhanced.

[0020] However, such simple signal combination methods rely on knowing the exact ratio of the combined input signals in order to filter out unwanted components, or they result in multiple possible output signals, and it is unclear which one should be used for further analysis.

[0021] Therefore, according to embodiments of the invention, alternatively, it is proposed to generate multiple candidate signals formed by different combination ratios of the input signals, and then apply another signal selection step for selecting the optimal candidate signal for a specific physiological phenomenon. This can be based on predetermined signal characteristics. Thus, the combination of signal fusion and then signal selection based on criteria specific to the physiological phenomenon in question allows for the extraction of signal components related to specific physiological sources in the body.

[0022] As described above, the system is configured to select from candidate signals based on signal feature analysis. In some examples, this may include a scoring procedure that scores candidate signals according to a determined probability that the signal represents the physiological source in question, and such that the signal with the highest score is selected for further analysis.

[0023] The system can be arranged to receive inputs corresponding to the sensed inductive signals from outside the system, and the system is configured to execute only a signal extraction procedure. For example, it may include a processor or controller unit for this purpose.

[0024] In other embodiments, the system may further include an inductive sensing device for acquiring inductive sensing signals.

[0025] In particular, according to one or more embodiments, the system further includes an inductive sensing device, the inductive sensing device comprising: A resonator circuit, which includes a loop antenna; A signal generation unit adapted to excite the loop antenna to generate the electromagnetic excitation signal; and A signal sensing unit adapted to use the loop antenna to sense the return signal from the body based on the detected change in the electrical characteristics of the resonator circuit.

[0026] In some examples, the system may include a processor, such as a microprocessor unit, which is configured to control the resonator circuit, signal generation unit, and signal sensing unit to perform the signal detection, combination, and selection steps outlined above.

[0027] The first and second input signals can represent the variation or change of frequency or amplitude over time. They can represent the deviation of the starting frequency or amplitude of the current in the self-resonant circuit. They can indicate the deviation of the natural (e.g., resonant) frequency and natural (e.g., resonant) amplitude of the self-resonant circuit. The amplitude signal can be a damped signal indicating the damping of the current over time (i.e., the change in the natural amplitude due to the return signal). It may also be referred to herein as an absorption signal because it indicates the absorption of energy in the applied excitation signal, resulting in a measurable change in the amplitude of the current in the resonant circuit.

[0028] According to one or more examples, the signal generation procedure is based on the use of the Independent Component Analysis (ICA) method.

[0029] For example, the ICA method can be used to determine the combination ratio used to generate the signal. ICA is a well-known signal analysis method and is based on the assumption that the returned signal sensed at the resonator circuit antenna is a composite signal formed by multiple signal components, each corresponding to a different physiological source.

[0030] ICA seeks to determine a weight vector matrix describing how potential physiological signal components are mapped to two detected input signals. This allows the original physiological signal components to be reconstructed from a specific linear combination of the input signals.

[0031] ICA procedures can be quite resource-intensive. Therefore, as an alternative, according to another set of embodiments, the signal generation procedure can form the plurality of candidate signals based on the use of a set of predefined signal combination ratios.

[0032] For example, signal combination ratios can be stored in a list. Alternatively, predefined signal protocols can exist, which define schemes for combining input signals at a range of different ratios. For instance, a series of combination ratios can be defined at set intervals.

[0033] Therefore, in some examples, this can be used in place of ICA (e.g., this can increase processing speed), or in some examples, it can be used in combination with ICA to improve the accuracy of signal selection.

[0034] According to one or more embodiments, the criteria for the signal selection procedure may include one or more of the following: the frequency of the candidate signal, and the number of maximum and minimum values ​​of the signal within a given time window.

[0035] According to one or more embodiments, the signal extraction procedure may further include generating information output indicating the specific physiological phenomenon based on selected candidate signals.

[0036] According to one or more embodiments, the signal extraction procedure may include an additional step of applying a bandpass filter to the input signal prior to the signal generation procedure.

[0037] For example, the threshold or parameters of the bandpass filter can be set according to the physiological phenomenon from which the signal is to be extracted. This can be based on, for example, predefined parameters known to be associated with different physiological phenomena. This advantageously suppresses signal frequency components known to be outside the range of signal frequency components associated with the physiological phenomenon in question.

[0038] According to one or more embodiments, the signal extraction procedure may include another signal processing step applied directly after the input signal is detected, the signal processing step being configured to suppress motion artifacts in each of the input signals.

[0039] In one set of advantageous embodiments, the other processing step may include: Receives the fundamental frequency input indicating the motion to be suppressed; and A notch filter is applied to the input signal, the notch filter having an adjustable frequency setting, wherein the notch filter is applied to the input signal at one or more times the fundamental frequency.

[0040] This program allows for the suppression of periodic motion artifacts that may be present in the input signal.

[0041] In some examples, the fundamental frequency can be determined based on input from a motion sensor, such as an accelerometer. It can be determined by the system itself or simply received from outside the system.

[0042] In an advantageous example, the selection procedure is configured in at least one mode to select candidate signals determined to indicate the respiratory rate of the object.

[0043] This can be based on, for example, predetermined signal characteristics known to be associated with respiratory signals.

[0044] The system can be configured in at least one mode to run two iterations of the signal extraction procedure, wherein the signal selection procedure is configured in the first and second runs to select signals associated with different corresponding first and second physiological phenomena.

[0045] In a preferred set of examples, at least in the second run, a bandpass filter may be applied to the sensed input signal prior to the signal generation procedure.

[0046] In a preferred set of examples, the signal selection procedure may be configured in the first run to select a signal related to the subject's respiratory rate, and in the second run to select a signal related to the subject's heart rate.

[0047] An example of another aspect of the invention provides a method for processing electromagnetic signals returned from a body in inductive sensing in response to an electromagnetic excitation signal applied to the body, the method comprising: Receive a signal input indicating the sensed return signal, the return signal corresponding to a signal sensed at the loop antenna of the resonator circuit based on a change in the electrical characteristics of the resonator circuit detected when the circuit is driven to generate the excitation signal; and Implementing the signal extraction procedure includes: A first input signal is detected from the sensed return signal, the first input signal being based on the frequency of the sensed return signal. A second input signal is detected from the sensed return signal, the second input signal being based on the sensing amplitude of the sensed return signal. The application signal generation program includes generating multiple candidate signals, each candidate signal being formed by different linear combinations of the first input signal and the second input signal, and A signal selection procedure is applied to select one of the candidate signals, the selection procedure being based on predefined criteria related to one or more signal characteristics of the input signal, the criteria being configured to isolate signals related to specific physiological sources in the body, the selected signal forming an output signal.

[0048] This method may be used only to perform signal processing, and wherein the generation and sensing of physical sensing signals are performed separately, outside the scope of the claimed method.

[0049] However, in another set of embodiments, the method may further include the following steps: An electromagnetic excitation signal is applied to the body using a resonator circuit, said resonator circuit including a loop antenna; and The loop antenna is used to sense the return signal from the body based on the detected change in the electrical characteristics of the resonator circuit.

[0050] Therefore, in this other set of embodiments, the method further includes a step for performing the physical sensing itself.

[0051] An example of another aspect of the invention provides a computer program product including code units configured to, when run on a processor, cause the processor to perform a method according to any example or embodiment outlined above or described below.

[0052] These and other aspects of the invention will become apparent with reference to one or more embodiments described below, and will be illustrated with reference to one or more embodiments described below. Attached Figure Description

[0053] To better understand the invention and to more clearly illustrate how it can be implemented, reference will now be made to the accompanying drawings by way of example only, wherein: Figure 1 The basic principle of inductive sensing is illustrated schematically. Figure 2 The components of an example inductive sensing system according to one or more embodiments are schematically outlined in block diagram form; Figure 3 The steps of an example signal extraction procedure applied by a system or method according to one or more embodiments are outlined; Figure 4 An example input signal sensed at an inductive sensor antenna is illustrated, as is an example candidate signal formed by a linear combination of the two. Figure 5 The illustration shows a set of example candidate signals for respiratory rate, each candidate signal being formed by a linear combination of at least two input signals; Figure 6 The illustration shows a set of example candidate signals for heart rate, each candidate signal being formed by a linear combination of at least two input signals; Figure 7 The steps of an example procedure for determining the frequency of motion artifacts are outlined. Figure 8 The illustration shows an example notch filter function used in motion suppression procedures; Figure 9 The steps in an example procedure for extracting respiratory rate measurements from the output signal of a signal extraction procedure are outlined. Figure 10 The steps of an example method for extracting heart rate measurements from the output signal of a signal extraction procedure are outlined. Figure 11 The illustration depicts the steps of an example signal extraction method according to one or more embodiments; and Figure 12 The illustration shows the steps of another example signal extraction method according to one or more embodiments. Detailed Implementation

[0054] The present invention will be described with reference to the accompanying drawings.

[0055] It should be understood that the detailed descriptions and specific examples, while indicating exemplary embodiments of the devices, systems, and methods, are intended for illustrative purposes only and are not intended to limit the scope of the invention. These and other features, aspects, and advantages of the devices, systems, and methods of the present invention will be better understood from the following description, the appended claims, and the accompanying drawings. It should be understood that the drawings are merely schematic and not drawn to scale. It should also be understood that the same reference numerals are used throughout the drawings to denote the same or similar parts.

[0056] This invention provides a system and method for extracting component signals related to different physiological phenomena in the body from sensed inductive signals. A resonator circuit oscillates at a certain frequency to generate an alternating electromagnetic field applied to the body under study. This field induces secondary eddy currents in the body, which interact with the primary magnetic field and at least alter the frequency and amplitude of the oscillating current in the resonator circuit. These changes in current characteristics (particularly frequency and amplitude) are measured, providing a first input signal and a second input signal. Embodiments of the invention provide a system or method arranged to receive these input signals. The system then generates multiple different composite or fused signals, each formed by different linear combination ratios of the two input signals. These are then evaluated using a signal selection procedure to identify the best candidate signals for providing a measure or indication of a specific physiological phenomenon or one or more. This can be based on, for example, predefined selection criteria related to the signal characteristics of the candidate signals.

[0057] Embodiments of the present invention provide systems and methods for processing sensing signals acquired by inductive sensing devices. In some cases, the systems and methods perform steps for acquiring inductive sensing signals. Therefore, the embodiments are generally based on the principle of magnetic induction. The basic principle of magnetic induction will first be briefly outlined.

[0058] Inductive sensing is based on the principle of inductive coupling, whereby a coil or wire has an induced potential difference due to exposure to a time-varying magnetic field. Embodiments of the present invention use this principle to measure the intensity of electromagnetic signals generated within a body region by sensing changes in the inductance of a coil or loop antenna placed close to the body, wherein these changes are detected based on alterations in the resonant characteristics of the antenna or resonator circuit.

[0059] Some embodiments of the present invention utilize a resonator, including an antenna (which in a preferred embodiment may consist of only a single-turn loop) to stimulate or excite the body with electromagnetic signals (waves) and to sense signals emitted back from the body in response to those excitation signals.

[0060] Coils can be driven by alternating current to generate excitation signals for application to the body. These can be propagating electromagnetic signals, which in some cases propagate into a medium, or the signals can include non-propagating electromagnetic fields applied to the medium, i.e., by bringing a loop antenna source close to the target medium. Alternating current produces a field with alternating field strength.

[0061] When the coil approaches the body, the inductor L collects the additional reflected inductance component due to the eddy currents induced in the body by the stimulation. L r As a result of the application of an excitation signal.

[0062] This is Figure 1 The diagram schematically illustrates, by way of example, an alternating current driven loop antenna 12 approaching the chest 16 of an object in order to apply an electromagnetic signal 22 to the chest cavity.

[0063] Therefore, vortex 18 is sensed within the thoracic cavity.

[0064] These eddy currents then effectively contribute to the inductance of the loop antenna 12. This is because they themselves cause a time-varying magnetic flux 24 at an equivalent frequency generated by the primary antenna 12. These eddy currents combine with the primary flux of the antenna, which results in a modified inductive anti-EMF in the antenna, and therefore a larger measurable effective inductance.

[0065] The added component of inductance caused by eddy currents is called "reflected inductance". L r The total inductance of coil antenna 12 L t It can be expressed as: L t = L 0 + L r in, L 0 It is the self-inductance of the coil antenna 12 in free space, and L r It is the reflected inductance caused by the presence of the body.

[0066] Typically, reflective inductors L r It is a complex number and can be expressed as: in, It is related to the reactive impedance of the coil antenna, and It is related to the resistance impedance of the coil.

[0067] inductance L r The addition of the reflected component causes detuning of the electrical characteristics of the resonator circuit. Specifically, both the natural radial frequency and the damping factor of the resonator circuit change. By measuring this detuning of the electrical characteristics, the reflected inductance can be detected. L r The real and imaginary parts.

[0068] Specifically, the additional inductance component L r The real part of the inductance component is represented in the frequency of the resonator circuit or antenna. The imaginary part of the additional inductance component is represented in the amplitude of the resonator circuit. Therefore, by measuring the changes in the frequency and amplitude of the resonator circuit (current) and deriving the first and second input signals respectively, signals indicating potential anatomical movements and phenomena can be detected.

[0069] For brevity and ease of description, embodiments of the present invention will be described below, which include components or method steps for both generating and sensing sensing signals and performing processing of the sensing signals (particularly performing a signal extraction procedure). However, it should be understood that embodiments of the present invention may include only components for performing signal processing, wherein the input sensing signal is received at the system as a signal input. For example, one set of embodiments may include only a processor or control unit configured to perform a signal extraction procedure. Therefore, the descriptions and options set forth below should be understood to apply equally to embodiments in which the system includes only the unit for performing the signal extraction procedure.

[0070] Figure 2 A block diagram of components of an example inductive sensing system 8 according to one or more embodiments is shown.

[0071] The inductive sensing system 8 is used to sense electromagnetic signals returned from the body in response to the application of an electromagnetic excitation signal to the body.

[0072] The system includes a resonator circuit 10, which comprises a loop antenna 12 and an electrically coupled capacitor 13. The capacitance of capacitor 13 at least partially defines the natural resonant frequency of the resonator circuit (in the absence of force or damping). When the antenna 12 is excited, it will tend to resonate naturally at the defined resonant frequency, generating an electromagnetic signal at the same frequency. Therefore, selecting the capacitance of the capacitor allows at least partial tuning of the frequency of the generated electromagnetic signal.

[0073] System 8 also includes a signal generation unit 14 adapted to excite the loop antenna to generate an electromagnetic excitation signal. The signal generation unit may include a driver unit for, for example, at a radial frequency... Driven antenna, that is, using frequency An AC-driven antenna. The driving unit may be, or includes, for example, an oscillator.

[0074] The signal generation unit can utilize radial frequency The current-driven antenna and resonator circuit, wherein the radial frequency is required to be... The excitation signal.

[0075] By exciting the resonator, a resonant current is induced to flow back and forth through the loop antenna into the capacitor. By driving alternating current through the antenna, an oscillating electromagnetic signal (wave) can be generated.

[0076] The same antenna used to generate the excitation signal is the same antenna that is used in response to sensing electromagnetic signals received from the body.

[0077] To avoid any doubt, “electromagnetic excitation signal” simply means an electromagnetic signal applied to the body for the purpose of stimulating or arousing the generation of eddies within the body, which is then used to stimulate electromagnetic signals emitted in the opposite direction to the body that can be sensed by a sensing system.

[0078] The term "electromagnetic signal" can generally refer to electromagnetic radiation emission, electromagnetic near-field oscillation, electromagnetic oscillation, and / or electromagnetic waves.

[0079] System 8 also includes a signal sensing unit (“signal sensing”) 20, which is adapted to sense the returned signal from the body using the loop antenna 12 based on changes in the electrical characteristics of the detected resonator circuit 10. The signal sensing unit may include a signal processing or analysis unit for detecting or monitoring the electrical characteristics of the current in the resonator circuit 10.

[0080] For example, the signal sensing unit 20 can be adapted to monitor at least the frequency of the resonator circuit current and the amplitude of the resonator circuit current. These properties of the current will vary depending on the strength of the reflected electromagnetic signal returning from the body and detected at the antenna.

[0081] The sensing of these signal characteristics is performed simultaneously with the excitation of the antenna used to generate the excitation signal. Therefore, signal transmission and sensing are performed concurrently.

[0082] The system is configured to implement a signal extraction procedure. For example, a processor or controller unit 56 may be provided to execute the signal extraction procedure. Figure 3 The steps of an example signal extraction procedure 30 according to one or more embodiments are summarized in the form of a block diagram.

[0083] The system is configured to detect a first input signal 32 from the sensed return signal, the first input signal being based on the frequency of the sensed return signal.

[0084] The system is also configured to detect a second input signal 34 from the sensed return signal, the second input signal being based on the sensed amplitude of the sensed return signal at the antenna.

[0085] The system is also configured to apply a 36-signal generation procedure, including generating multiple candidate signals, each candidate signal being formed by a different linear combination of a first input signal and a second input signal.

[0086] The system is also configured to apply a signal selection procedure 38 for selecting one of the candidate signals. The selection procedure is based on predefined criteria associated with one or more signal characteristics of the input signal, which are configured to isolate signals related to specific physiological sources in the body. The selected signal forms the output signal. For example, according to one or more application embodiments, the selection criteria can be configured to select signals related to the subject's respiratory rate and / or heart rate signals.

[0087] System 8 is adapted to perform the series of steps outlined above using the components described above. The system may include a controller or microprocessor (“MPU”) 56 adapted to perform or facilitate these steps. Example microprocessor 56 is shown in... Figure 2 The example system is shown for illustration purposes. However, a dedicated controller or microprocessor is not required. In other examples, one or more other components of the system, such as the signal sensing unit 20 and / or the signal generation unit 14, may be adapted to perform the steps.

[0088] exist Figure 2 In the diagram, the signal sensing unit 20 is shown connected to the resonator circuit 10 via the signal generation unit 14. However, this is not necessary: ​​the signal sensing unit and the signal processing unit can be connected to the resonator independently.

[0089] In some examples, the signal extraction procedure can be configured to extract respiratory signals, such as respiratory rate. In other examples, it can be configured to extract heart rate. These represent only two favorable examples.

[0090] As described above, according to one or more embodiments, the resonator circuit 10, signal generation unit 14, and signal sensing unit 20 can be omitted from system 8. The system may consist only of a processor unit 56 configured to receive a signal input indicating a return signal sensed from the body, the return signal corresponding to a signal sensed at the loop antenna of the resonator circuit, for example based on a change in the electrical characteristics of the resonator circuit detected when the circuit is driven to generate an excitation signal.

[0091] In a preferred set of examples, the system is configured to run at least two iterations or runs of the signal extraction method, the first for extracting a signal related to the subject’s respiratory rate and the second for extracting a signal indicating heart rate (or vice versa).

[0092] The signal generation process involves fusing or linearly combining input signals at different ratios. This means adding or superimposing input signals with different linear combination coefficients. For example, a set of sample candidate signals C generated from input signals s1 and s2 through this process may include: C1 = 1.0 s1 + 1.0 s2 C2 = 1.5 s1 + 1.0 s2 C3 = 1.0 s1 + 1.5 s2 … C n = α s1 + β s2 α and β can be positive or negative.

[0093] To further explain, the inductive sensing device according to an embodiment of the present invention measures changes in the volume of the human body. However, the input signal sensed at the antenna typically contains components of various physiological phenomena. Furthermore, some phenomena are stronger in the sensed signal than others. For example, the respiratory signal component is stronger than the heart signal component, and therefore, for an antenna placed on the chest, the respiratory signal component can dominate the input signal, making it difficult to detect the heart rate signal.

[0094] Similarly, the presence of a cardiac component is problematic for respiratory rate measurements. While respiratory rates can reach as high as 60 breaths per minute, heart rates can be as low as 30 heartbeats per minute. Therefore, both phenomena share a frequency range between 30 and 60 BPM, meaning that purely frequency-based signal separation is not feasible.

[0095] Signal fusion is a method in which signals from various sources are combined to eliminate unwanted components. Since the contributions of volume changes caused by respiration and heartbeat to frequency and absorption are different, if signals are linearly combined (added together or subtracted from each other by a factor), the resulting signal is, in most cases, better suited for extracting one of these physiological signals (heart rate and respiratory rate) than the other, and more often for extracting respiratory rate than heart rate. By combining signals at an appropriate linear multiple or ratio, signal components relevant to the desired physiological signal can be emphasized while those from other physiological phenomena can be suppressed.

[0096] This is Figure 4 The diagrams are illustrated schematically. For reference, signal (a) shows a real respiratory signal, for example, measured by an additional auxiliary sensor. Signal (b) illustrates an example first input signal (indicating the frequency of the received induced signal sensed at antenna 12). Signal (c) illustrates an example second input signal (indicating the body's energy absorption, for example, the amplitude of the induced signal sensed at antenna 12). Signal (d) illustrates an example candidate signal formed by a linear combination of input signals (a) and (b) (particularly formed by the difference between the frequency signal (signal (b)) and the absorption signal (signal (c))).

[0097] The signal generation procedure 36 can be performed in different ways. Specifically, a common approach is to determine or calculate a specific combination ratio or set of ratios of the input signal to maximize the emphasis of the desired physiological signal component, and then select the best signal from this selection set. Another common approach is to simply generate a large number of candidate signals formed by different mixing ratios of the input signal, where the mixing ratios are random or follow a criterion sequence of incrementally varying combination coefficients, and then determine the best signal from this larger set. Examples of these two methods will now be outlined.

[0098] Various methods exist for searching specific mixing matrices for multichannel signals that separate certain features contained within the signal. In a particular example, each input signal can be considered a multichannel signal (i.e., containing signal components from multiple physiological sources). A mixing matrix is ​​a set of linear combination coefficients used to combine multichannel input signals to emphasize or weaken certain signal components.

[0099] Based on one or more examples, the Independent Component Analysis (ICA) procedure can be applied to determine the combination ratio or linear combination coefficients of the different input signals used to form multiple candidate signals.

[0100] ICA is a well-known signal analysis method, and it is based on the following assumption: the return signal sensed at the resonator circuit antenna is a composite signal formed by multiple signal components, each of which corresponds to a different physiological source.

[0101] ICA seeks to determine a weight vector matrix describing how potential physiological signal components are mapped to two detected input signals. This allows the original physiological signal components to be reconstructed from a specific linear combination of the input signals.

[0102] The multi-channel signal (i.e., the input signal) is decomposed into independent non-Gaussian signals using ICA (Independent Component Analysis). Based on this method, the induced frequency and absorption can be considered as two axes of a 2D vector array.

[0103] Generally speaking, ICA can be understood as a rotating signal vector that makes each axis appear as non-Gaussian as possible.

[0104] Based on one or more examples, the first step could be to whiten (or “spherize”) the data (i.e., the input signal). This makes the mean zero (centered) and normalizes the variance in all directions (whitening), thus effectively containing the vector within a sphere near zero.

[0105] A popular approach to ICA suitable for use in signal generation procedures based on one or more examples is called Fast ICA. The details of this algorithm are described, for example, in the paper [Hyvärinen, A., & Oja, E. (2000). Independentcomponent analysis: algorithms and applications. Neural Networks, 13, 411–430.]. It uses an efficient approximation of the entropy of the potential source signal and has an efficient iterative method for optimizing the weight vector matrix (or its inverse) to minimize the entropy (maximize non-Gaussianity).

[0106] Based on one or more examples, instead of using ICA to generate candidate signals, signals can be combined randomly or with a standard predefined sequence or a set of combinations (combination ratios or coefficients).

[0107] For example, candidate signals can be generated according to a predefined signal generation protocol that defines sequentially increasing and / or decreasing linear combination coefficients for two or more input signals. For instance, the signal generation protocol can efficiently stepwise through a sequence of increasing or decreasing linear combination coefficients for each input signal with regular increments, thereby generating a set of candidate signals that follow a standard sequence of increasing linear combination coefficients.

[0108] For example, let r This represents the combination ratio, with a uniform step size of 0.1: r = [0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0] This can be achieved by inducing a frequency signal (the first input signal). I f and induced absorption signal I a (The second input signal) is added and subtracted to generate two series of signal combinations: s 1 = r ∙ I f+ (1 – r) ∙ I a s 2 = r ∙ I f – (1 – r) ∙ I a Note that the final sign (i.e., positive or negative) and final scale (i.e., amplitude) of the combined (candidate) signals are not critical. What matters is the relative combination ratio of the input signals included in the candidate signals.

[0109] One possible improvement to the procedure could be, for example, using an adaptive step size (i.e., a non-uniform step size) based on gradient descent search.

[0110] In cases where only a small number of input signals exist (e.g., only two input signals), as in this example, the computational load of generating 100 different candidate signals according to a predefined signal generation protocol (such as the one discussed above) is (most of the time) faster than a single ICA calculation. Therefore, based on one or more examples, the computational requirements or usage can be advantageously reduced by following a random signal generation process or by following a fixed protocol or a set of linearly combined coefficients instead of using an ICA procedure.

[0111] Once candidate signals have been generated, it is necessary to determine which one (or more) to use for extracting the measurement or sample of the physiological parameter or signal in question (e.g., respiratory rate or heart rate based on the two examples). Therefore, a signal selection procedure is applied.

[0112] According to one or more examples, the signal selection procedure may include a scoring procedure or algorithm in which each candidate signal is “scored” based on the determined probability that it best reflects the physiological parameter in question (e.g., based on the degree to which the signal is “similar to respiratory rate” or the degree to which the signal is “similar to heart rate”).

[0113] Scoring can be based on, for example, one or more predefined signal characteristics known to be associated with signals reflecting a particular physiological phenomenon or phenomenon. In other words, scoring can be accomplished by looking at the typical characteristics for the physiological phenomenon in question.

[0114] Even when using ICA to generate initial candidate signals, using a signal selection procedure is advantageous. This is because, while ICA can successfully separate one or more specific physiological signals from other distortions or other physiological signals (e.g., cardiac or motion signals), it cannot identify which of the resulting separated signal components corresponds to a specific physiological phenomenon. In other words, it cannot attribute the corresponding source to the different separated signal components; it cannot identify which physiological phenomenon or noise source each separated signal component belongs to.

[0115] Therefore, additional steps are needed to identify which of the generated candidate signals is related to the specific physiological phenomenon of the desired output signal. For example, additional steps are needed to “score” each output signal based on the degree to which each output signal possesses signal characteristics that match those known to be associated with the physiological signal source in question.

[0116] For example, to identify the best candidate for representing a subject's respiratory rate, one possible approach is to analyze how "low-frequency" the signal is. The typical frequency of respiratory rate is below both the typical frequency range of the heart pulse and the typical frequency range of periodic motion distortion caused by walking.

[0117] In this example, scoring can be accomplished, for example, by counting the maximum and minimum values ​​in the signal. The signal with the lowest count receives the highest score and is selected as the output signal.

[0118] This method is a simple and fast way to perform signal scoring. It is, for example, simpler than using a bandpass filter. A bandpass filter alone cannot determine whether a signal is low-frequency. To determine this, it would be necessary to apply multiple bandpass filters and identify which of the resulting filtered signals has the highest energy, or to perform a Fourier transform and analyze the spectrum. These are expensive operations in terms of processing resources. Counting the zero-crossings of peaks / valleys is a simpler and faster alternative.

[0119] This is Figure 5 The diagram is shown in the image. Figure 5 A set of four example candidate signals is depicted, and the number of peaks and troughs, i.e., scores, are listed for each candidate signal. The “best” signal is selected as the signal with the lowest number of peaks and troughs, i.e., signal-1.

[0120] When using ICA, as mentioned above, the scoring can be applied to only the two output signals of the ICA program, or in addition to the two original input signals of the ICA program (from the ICA formula above). I f and I aIn addition to the above, it can also be applied to both outputs. In the latter case, this means that if the ICA procedure is unsuccessful in generating a better breathing-like signal, it will not affect the final result, as the original input signal can be chosen as a candidate signal instead.

[0121] As another example, to identify the best candidate for indicating heart rate signals, according to one or more examples, a scoring of the candidate fused signals for heart rate measurement can be performed based on two signal characteristics: 1. For example, the frequency of a signal indicated by the number of zero-crossing points (Z). 2. Flatness (F) of the energy envelope. This is defined as the ratio between the (uncorrected) standard deviation (S) and the mean absolute deviation (M) multiplied by a factor of 200 (e.g., determined empirically).

[0122] Different candidate signals can be scored based on their corresponding Z and F values, where higher scores are attributed to higher Z and F values.

[0123] This method is based on the understanding that non-cardiac-related signal components typically exhibit lower frequencies than heart rate signals (respiratory signals), and that the signal envelope of cardiac pulses (within a time window) is generally relatively constant.

[0124] The values ​​of two features can be summed to form the final score: score = Z + F.

[0125] The example method is in Figure 6 The diagram is shown in the image. Figure 6 The diagram illustrates a set of four example candidate signals, along with the corresponding Z and F values ​​for each candidate signal. The "best" candidate signal (i.e., the candidate signal with the highest Z and F values) is indicated. In this example, the selected signal is signal -1.

[0126] According to one or more embodiments, the signal extraction procedure executed by the system may include additional signal processing steps applied directly after the detection of input signals 32, 34, which are configured to suppress motion artifacts in each input signal.

[0127] Motion suppression can be accomplished in different ways. Depending on one or more examples, it can involve the simple application of filters (e.g., bandpass filters). Filter parameters can be configured to select frequency components that are known to be typical of the physiological phenomenon or parameter in question, or at least to exclude one or more frequency components that are known to be outside the typical range of the phenomenon in question.

[0128] A favorable method for suppressing motion artifacts in an input signal will now be briefly outlined. This method is particularly suitable for filtering or removing periodic motion artifacts (i.e., those exhibiting a periodic frequency). Such artifacts can be caused, for example, by a patient walking or running (where the sensor is a portable wearable sensor) or by other regular movements affecting the object.

[0129] In this example, an additional motion suppression processing step is based on detecting or otherwise identifying the fundamental frequency of the periodic motion to be suppressed, and then applying a filter to remove frequency components around the range of that fundamental frequency.

[0130] For example, the system can be configured to receive an input of a fundamental frequency indicating the motion to be suppressed, and to apply a filter (notch filter) to the input signal, the notch filter having an adjustable frequency setting, wherein the notch filter is applied to the input signal at one or more times the fundamental frequency.

[0131] As an example, motion artifacts associated with walking and running are typically confined to certain frequency bands in the inductive sensor signal. Therefore, these periodic distortions can be suppressed, for example, by using an adaptive notch filter at multiples of the fundamental frequency (f0) of the periodic motion.

[0132] The fundamental frequency f0 of the periodic motion can be determined or detected by the system itself, or it can be received as an input to the system, where it has already been determined or identified externally. It can be detected, for example, by an external motion sensor element and transmitted to the system of the present invention.

[0133] According to one or more examples, the motion frequency can be determined by the system based on input motion signals from, for example, motion sensors carried by the object.

[0134] For example, input from an accelerometer (e.g., a 3D accelerometer coupled to an object) can be used to measure motion frequency.

[0135] A 3D accelerometer measures acceleration in three directions: X, Y, and Z (in the accelerometer's frame of reference). These can be referred to as the body axes.

[0136] A 3D accelerometer measures both acceleration caused by gravity and acceleration caused by motion. Gravity introduces a constant bias (g) distributed across the axes. To extract only the acceleration caused by motion... The gravitational component must be subtracted. This can be done in two steps: First, the norm of the acceleration vector can be calculated: Secondly, the gravity deviation can be removed by subtracting the constant gravity, and the signal is corrected: in, The motion suppression process may include an initial step of detecting whether one or more input signals contain motion artifacts that need to be suppressed. The remaining steps of the motion suppression process can then be applied only if motion artifacts are detected.

[0137] As an example, the motion threshold T can be... m Applied to signal A m And among them, those exceeding the threshold T m motion signal A m (For example, after windowing using a 10-second Hanning window) is considered an indication that the signal has motion artifacts that need to be suppressed.

[0138] The threshold can be predefined and stored locally in the system, or it can be determined empirically, for example. As an example, in one sample application, a volunteer study was conducted with 15 volunteers, each wearing a combination of an inductive sensing device and an accelerometer according to an embodiment. Measurements were repeated at a frequency of once per second. As a result of this study, 1 m / s was identified. 2 Threshold T m In some embodiments, 1 m / s can be used. 2 Threshold T m .

[0139] As mentioned above, the motion frequency can be measured using signals from a 3D accelerometer. Specifically, the frequency can be determined based on the norm of the 3D accelerometer signal. The distortion introduced by the left and right feet may differ slightly. In some examples, the fundamental frequency f0 measured from the acceleration can therefore be divided by 2 to suppress these distortions.

[0140] By examining the measured spectra of the acceleration frequency and the sensed induced signal frequency separately, it has been found that the main contribution to motion distortion is at about half of the motion fundamental frequency f0, on the order of about 50 rpm.

[0141] Many methods for extracting the fundamental frequency f0 from an input signal are known in the art, and those skilled in the art will recognize such exemplary methods. As an example, fundamental frequency detection is typically performed based on finding a peak in the amplitude spectrum of the signal or in a time-domain correlation function (such as autocorrelation) of the signal.

[0142] In a preferred embodiment, a particularly advantageous algorithm known as combined spectral pitch detection can be used to determine the fundamental frequency f0 from the input acceleration signal. The full details of this algorithm are described in document WO2012 / 063185, which readers can refer to for further implementation details.

[0143] In short, typical methods of detecting the fundamental frequency based on the peak value in the detection amplitude or time-domain correlation function can lead to false detections at multiples of the fundamental frequency. For spectral analysis, this might be, for example, higher harmonics. For time-domain correlation, it could be multiples of the repetitive pulse signal.

[0144] Alternative pitch detection methods (described in WO2012 / 063185) are based on combining frequency and time domain signals so that the resulting signal has only the f0 component.

[0145] Figure 7 The steps of the combined spectral pitch detection algorithm are briefly outlined in the form of a block diagram.

[0146] In the first step 62, the input signal S is windowed by applying a window function. Further details can be found on page 8, lines 14-30 of WO2012 / 063185.

[0147] Then, in step 64, the obtained windowed signal S is processed based on the application of the Discrete Fourier Transform (DFT). w The signal is transformed from the time domain to the frequency domain to provide its spectrum. For efficiency reasons, a Fast Fourier Transform (FFT) (e.g., radix-2 FFT) is preferred. Further details can be found on page 8, line 31 to page 9, line 9 of WO2012 / 063185.

[0148] Extract the amplitude of the signal |S|. Further details can be found on page 9, lines 10 to 25 of WO2012 / 063185.

[0149] Then, in step 66, the window-compressed amplitude spectrum is transformed to the time domain using the inverse Fourier transform (IFT). For example, the inverse fast Fourier transform (IFFT) can be used. This time-domain transformation is used to obtain the correlation signal c, which includes peaks at multiples of the fundamental frequency. Further details can be found on page 10, line 21 to page 11, line 9 of WO2012 / 063185.

[0150] In step 68, a combined spectrum b is formed by multiplying the amplitude spectrum S with the associated signal c. This combined spectrum b has a distinct peak at the fundamental frequency. By multiplying these spectra, higher harmonics in the spectrum are attenuated, and the fundamental frequency remains the dominant peak. Further details can be found on page 11, line 18 to page 12, line 25 of WO2012 / 063185.

[0151] Finally, in step 70, a peak position detection step is performed, which includes searching for the maximum value of the combined spectrum b. This results in an output frequency value p, in Hz, indicating the fundamental frequency f0. Further details can be found on page 13, line 1 to line 13 of WO2012 / 063185.

[0152] As described above, once the fundamental frequency f0 of the motion artifact has been detected or received by the system, the motion artifact in the input signal can be suppressed by applying a filter with frequency parameters set according to the fundamental frequency. For example, the filter can be a notch filter (e.g., an adaptive notch filter), which allows the frequency parameters to be adjusted and multiple frequency components to be filtered out.

[0153] By way of example, when the physiological phenomenon to be measured is respiratory rate, both the first and second input signals (frequency signal and amplitude or absorption signal) can be filtered by a notch filter at f0 and f0 / 2Hz, where f0 is the fundamental frequency of the periodic motion artifact.

[0154] Using diagrams Figure 8 The diagram illustrates the function of a notch filter applied at a frequency of 40 rpm, with various quality factors Q ranging from Q=2 to Q=6. Lines 102-110 show the filter functions for Q factors of 2, 3, 4, 5, and 6, respectively.

[0155] By way of example, when the physiological phenomenon to be extracted is respiratory rate, Q can advantageously be set to Q=5. When the physiological phenomenon to be extracted is heart rate, for heart rate measurement, Q can advantageously be set to Q=6.

[0156] For heart rate calculation, it is preferable to suppress only f0 / 2Hz. This is because the fundamental frequency f0 of the associated motion artifacts is typically much higher than the respiratory rate, while for heart rate measurement, motion f0 is usually closer to the heart rate frequency. Therefore, applying a filter at f0 in the case of heart rate detection could result in the complete removal of the fundamental frequency of the cardiac pulse signal itself. Therefore, to avoid this, it is preferable that the notch filter is set only at f0 / 2Hz.

[0157] Once the best candidate signal has been selected in the signal selection process (as described above) Figure 3If step 38 of the method is followed, further steps may be required to actually extract the measurement of the physiological phenomenon in question from the selected candidate signal.

[0158] In some examples, this can be very simple and can be done, for example, by simply detecting the frequency of the candidate signal (e.g., to acquire respiratory rate or heart rate).

[0159] In other examples, additional processing steps can be applied to extract more accurate or precise measurements of physiological phenomena.

[0160] For example, when the physiological phenomenon or parameter to be acquired is a rate (such as respiratory rate or heart rate), the measurement of extracting the physiological parameter from the candidate signal may include detecting the fundamental frequency f0 from the selected candidate signal. This can help ensure that the derived measurement of the parameter or phenomenon is not unduly affected by noise-related frequency components in the signal.

[0161] As an example, we will now discuss extracting respiratory rate measurements or signals from selected candidate signals.

[0162] Figure 9 The steps of an example algorithm for extracting or deriving respiratory rate from selected candidate signals are outlined.

[0163] The program is based on detecting the fundamental frequency of the selected candidate signal s. This process is essentially the same as... Figure 7 The example algorithm described above is the same as the one presented in the previous section. Therefore, similar steps will not be described in detail here, but the reader can instead refer to the previous section on... Figure 7 Description of the relevant steps provided.

[0164] for Figure 7 The program begins with a windowing step 82, in which the input candidate signal s is time-windowed to generate a windowed signal S. w .

[0165] Then, in step 84, the obtained windowed signal S is processed based on the application of the Discrete Fourier Transform (DFT). w Transform the signal from the time domain to the frequency domain to provide its spectrum. For efficiency reasons, the Fast Fourier Transform (FFT) (e.g., radix-2 FFT) is preferred.

[0166] Extract the amplitude |S| of the signal. Figure 6Compared to the previous method, the extracted amplitude is increased to a power of 1.7. This increased exponent will have the effect of emphasizing spectral peaks and reducing noise contribution. The power value (1.7) was determined by the inventors empirically to be a particularly advantageous value. In the case of respiratory signals, this will (typically) have the effect of further emphasizing the fundamental frequency component (f0). For motion detection and heart rate extraction, this emphasis is not made because the fundamental frequency in these signals may be weaker than higher harmonics.

[0167] Then, in step 86, the window-compressed amplitude spectrum is transformed to the time domain using the inverse Fourier transform (IFT). For example, the inverse fast Fourier transform (IFFT) can be used. This transformation to the time domain is used to obtain the correlation signal c, which includes peaks at multiples of the fundamental frequency.

[0168] In step 88, a combined spectrum b is formed by multiplying the amplitude spectrum S with the associated signal c. This combined spectrum b has a distinct peak at the fundamental frequency. By multiplying these spectra, higher harmonics in the spectrum are attenuated, and the fundamental frequency remains the dominant peak. Further details can be found on page 11, line 18 to page 12, line 25 of WO2012 / 063185.

[0169] In step 90, a peak position detection step is performed, which includes searching for the maximum value of the combined spectrum b. This results in an output frequency value f0, in Hz, indicating the fundamental frequency of the respiratory rate. Further details can be found on page 13, line 1 to line 13 of WO2012 / 063185.

[0170] The calculated f0 can be used as a final respiratory rate measurement. In some examples, the algorithm parameters can be configured to provide measurements per second, each measurement within a 25-second window (therefore there is a 24-second overlap between windows). Each one-second measurement with a 25-second window is called a frame.

[0171] The method may optionally include an additional averaging step 94, wherein the output over multiple frames may be averaged over a longer duration to provide a more accurate prediction. For example, the derived frequency rate values ​​may be averaged over a one-minute window, as currently required by the World Health Organization (WHO) guidelines.

[0172] exist Figure 10 The following is a block diagram outlining a sample procedure for extracting heart rate from selected candidate signals s. The procedure is based on identifying the fundamental frequency f0 of the selected candidate signals s.

[0173] The example algorithm for determining the fundamental frequency (f0) of heart rate is similar to the description above regarding motion detection. Figure 7 The example algorithms outlined above are essentially the same. Therefore, similar steps will not be described in detail here; instead, the reader can refer to the section above. Figure 7 Description of the relevant steps provided.

[0174] The calculated f0 can be used as the final heart rate measurement. In some examples, the algorithm parameters can be configured to provide measurements per second, each measurement within a 25-second window (therefore there is a 24-second overlap between windows). Each one-second measurement with a 25-second window is called a frame. This example will have a corresponding (computational) delay of 12.5 seconds.

[0175] The method may optionally include an additional averaging step, wherein the output over multiple frames may be averaged over a longer period of time to provide more accurate predictions.

[0176] As described above, according to one or more embodiments, the signal extraction procedure may include an additional step of applying a bandpass filter to the input signal prior to the signal generation procedure. The bandpass filter allows the filtering out of any frequency components known to be outside the typical range of the physiological phenomenon in question, which makes the final extracted measurements more accurate and reliable.

[0177] The bandpass filter frequency can be set according to the physiological signal to be extracted.

[0178] As an example, when the physiological parameter to be extracted is respiratory rate, the bandpass filter can be set based on the typical range of respiratory rates of the target population.

[0179] Typical respiratory rates for adults range from 4 BPM (breaths per minute) to 60 BPM.

[0180] Therefore, bandpass filters can be configured to remove frequency components outside this range.

[0181] As an example, the filter may include, for instance, a cascade of two Butterworth IIR filters, wherein the second-order high-pass filter and the third-order low-pass filter have, for example, an attenuation of -6 dB at the cutoff frequency.

[0182] As an example, the analysis window duration can be set to 25 seconds. Filters can be applied individually for each analysis window in both the forward and reverse (zero-phase) directions without introducing filter delay.

[0183] As another example, when the physiological parameter to be extracted is heart rate, the bandpass filter can be set based on the typical range of heart rate or pulse rate of the target population.

[0184] The typical heart rate range for adults is from 30 BPM (heart beats per minute) to 220 BPM. Therefore, a bandpass filter can be set to remove frequency components outside this range.

[0185] In some examples, the latter can be used if the lower limit cutoff of 30 BPM is less than twice the respiratory rate.

[0186] The applied filter can be similar to the filter used for respiratory rate measurement. For example, it can consist of a cascade of two Butterworth IIR filters, where the second-order high-pass filter and the third-order low-pass filter have attenuation of, for example, -6 dB at the cutoff frequency.

[0187] The analysis window lasts for 25 seconds. Filters are applied separately for each analysis window in both the forward and reverse (zero phase) directions without introducing filter delay.

[0188] According to a set of advantageous embodiments, the system can be configured in at least one mode to perform at least two runs of a signal extraction procedure, wherein a signal selection procedure in the first and second runs is configured to select signals associated with different corresponding first and second physiological phenomena.

[0189] In a preferred example, in at least a second run, a bandpass filter is applied to the sensed input signal prior to the signal generation procedure.

[0190] The signal selection procedure may preferably be configured in a first run to select a signal related to the subject's respiratory rate, and in a second run to select a signal related to the subject's heart rate.

[0191] exist Figure 11 The example signal extraction method based on this approach is outlined in the form of a block diagram.

[0192] As described above, the method includes applying an electromagnetic excitation signal to the body of an object using a resonator circuit, which includes a loop antenna.

[0193] The method also includes using a loop antenna to sense the return signal from the body based on the detected change in the electrical characteristics of the resonator circuit.

[0194] From the sensed return signal, at least 32 first input signals and 34 second input signals are detected, the first input signal being based on the frequency of the sensed return signal and the second input signal being based on the sensed amplitude of the sensed return signal. The amplitude signal indicates the body's absorption of the initially applied stimulus signal.

[0195] Preferably, while detecting the input signal, the method may also include, in step 52, detecting the motion of the object (e.g., walking motion) using a motion sensor carried by the object (e.g., worn or attached to the person). The motion sensor may include, for example, an accelerometer, such as a 3D accelerometer.

[0196] If motion is detected, the system is configured to apply a motion suppression procedure in step 54 to suppress any motion artifacts in the detected input signal. (See above reference.) Figure 7 A sample motion suppression procedure is described in detail, and readers can refer to this description for details on the motion suppression procedure for step 54.

[0197] The method then continues to the first run or iteration for the step of extracting physiological signals or measurements from the input signal.

[0198] In the first run, the method was configured to extract a measurement of the subject's respiratory rate.

[0199] In the first run, the method preferably includes applying a 55-bandpass filter to each input signal. The frequency parameters of the bandpass filter are set, for example, based on a typical frequency range of the respiratory rate.

[0200] The method then includes applying a 36-signal generation procedure, which involves generating multiple candidate signals, each formed by a different linear combination of a first input signal and a second input signal. In other words, the induced frequency and the absorbed signal (i.e., the first input signal and the second input signal) are fused into each of a set of candidate signals. This can have the effect of, for example, obtaining a signal without any high-frequency distortion.

[0201] Then the respiratory rate is calculated.

[0202] This includes first applying a signal selection procedure 38 to select one of the generated candidate signals. This selection procedure is based on predefined criteria associated with one or more signal characteristics of the input signal, which are configured to isolate signals related to respiratory activity in the body. The selected signal forms the output signal. The criteria can be predetermined or predefined and can be based, for example, on empirical observations of typical signal characteristics that tend to reliably correlate with signals originating from respiratory movements in the body.

[0203] The selected candidate signal is used to calculate the 40 respiratory rate. The procedure for extracting the respiratory rate from the selected candidate signal has been described in detail previously in this disclosure.

[0204] The method then proceeds to a second run of the signal extraction procedure, this time configured to extract a measurement or signal of the heart rate of the indicated object.

[0205] After bandpass filtering 55, signal generation procedure 36 is applied again, including generating multiple candidate signals, each candidate signal being formed by different linear combinations or fusions of two input signals.

[0206] The heart rate is then calculated based on the reapplication of signal selection procedure 38, where, in this case, the selection criteria of the procedure are configured to select signals associated with the heart rate. Again, these criteria can be predetermined or predefined, for example, based on empirical studies or observations.

[0207] Once a candidate signal for calculating heart rate has been selected, 40 heart rate measurements are extracted from the selected signal. A sample procedure for extracting heart rate measurements from the selected candidate signal has been described in detail above.

[0208] The final result of this method is a set of output measurements that indicate respiratory rate and heart rate, respectively.

[0209] exist Figure 12 Another example signal extraction method based on this approach is outlined in the form of a block diagram, which is similar to... Figure 11 The example signal extraction method shown is provided, but additional feedback is provided from rate calculations 38, 40 to bandpass filtering 55.

[0210] The effective heart rate range (for adults) is approximately from 30 BPM to 220 BPM. When using biometric signals to measure heart rate, it is advantageous to remove any frequency components from the signal that fall outside the effective range, or at least to limit the search space to the minimum and maximum rates. Bandpass filters can be used to remove these unwanted frequency components. Any means of further limiting the effective range will improve measurement results.

[0211] according to Figure 12 The illustrated embodiment assumes that if the respiratory rate is known, it can be used to further limit the effective heart rate range. Limiting the heart rate range is even more advantageous if the biometric signal used to derive the heart rate also contains a frequency component related to the respiratory rate.

[0212] For example, an adult's respiratory range typically spans 4 to 60 breaths per minute. Therefore, based on a heart rate of 30 to 60 beats per minute, the measured fundamental frequency (f0) could belong to both heart rate and respiratory rate. Using the knowledge that heart rate is at least a first factor F1 (e.g., twice) of respiratory rate will help suppress one or more first harmonics (e.g., the first two harmonics) of respiratory rate in the signal, which will improve the accuracy of heart rate measurements.

[0213] Similarly, using the knowledge that the maximum heart rate is at a second factor F2 (e.g., ten times) of the respiratory rate to limit the upper limit of the effective heart rate will also improve the measurement results.

[0214] Therefore, in Figure 12In the illustrated embodiment, the output measurement 42 of the respiratory rate is fed back to a bandpass filter 55, which uses this knowledge to filter out frequency components in the detected input signal that cannot be part of the heart rate signal in order to determine the heart rate signal. For example, if the output measurement 42 of the respiratory rate is R1, the value of R1 is multiplied by a first factor F1 to give a lower limit of the frequencies in the detected input signal considered for determining the heart rate. Optionally, furthermore, the value of R1 is multiplied by a second factor F2 to give an upper limit of the frequencies in the detected input signal considered for determining the heart rate.

[0215] Factors F1 and F2 can be fixed values ​​(e.g., F1=2 and F2=10 for adults), or they can be preset or individually set by the user, for example, based on patient characteristics (e.g., age, sex, health status, etc.). The factors 2 and / or 10 applied to adults do not need to be exact. Values ​​close to 2 and / or 10, or other individual values, can also be used.

[0216] This embodiment can preferably be applied when respiratory rate and heart rate measurements are performed simultaneously. Respiratory rate measurement typically precedes heart rate measurement. In an exemplary embodiment, the output of the respiratory rate measurement is multiplied by 2, and if this value is higher than the system's expected minimum heart rate (e.g., 30 BPM), then twice the respiratory rate is used. As a further improvement, the upper limit of the heart rate range can also be limited to a maximum of 10 times the respiratory rate.

[0217] The implementation options and details of each of the above steps can be understood and interpreted based on the explanations and descriptions previously provided in this disclosure regarding the device aspect (i.e., the system aspect) of the invention.

[0218] Any examples, options, or embodiment features or details described above with respect to the device aspect (regarding the system) of the present invention may be applied, combined, or incorporated into this method aspect of the present invention with the necessary modifications.

[0219] Although the examples above have specifically referenced the extraction of respiratory rate and heart rate, these represent only two possible examples. Embodiments of the invention can be applied to extract any physiological or anatomical signal from the body, particularly those associated with or caused by movement of one or more body parts or features within the body. Inductive sensing is particularly suitable for detecting movement of a hydrated body.

[0220] An example of another aspect of the invention provides an inductive sensing method based on sensing an electromagnetic signal returned from the body in response to an electromagnetic excitation signal applied to the body.

[0221] In one set of embodiments, the method includes receiving a signal input indicating the sensed return signal, which corresponds to a signal sensed at the loop antenna of the resonator circuit based on a change in the electrical characteristics of the resonator circuit detected when the circuit is driven to generate an excitation signal.

[0222] The method also includes implementing a signal extraction procedure 30.

[0223] The signal extraction procedure includes detecting a first input signal from the sensed return signal 32, the first input signal being based on the frequency of the sensed return signal.

[0224] The signal extraction procedure also includes detecting a second input signal from the sensed return signal 34, the second input signal being based on the sensing amplitude of the sensed return signal.

[0225] The signal extraction procedure 30 also includes an application 36 signal generation procedure, which includes generating multiple candidate signals, each candidate signal being formed by a different linear combination of a first input signal and a second input signal.

[0226] The signal extraction procedure also includes a signal selection procedure, application 36, for selecting one of the candidate signals. This selection procedure is based on predefined criteria related to one or more signal characteristics of the input signal, which are configured to isolate signals associated with specific physiological sources in the body. The selected signal forms the output signal.

[0227] According to another set of embodiments, the method may further include steps for performing physical sensing and acquiring inductive sensing signals.

[0228] In particular, according to one or more embodiments, the method may further include: An electromagnetic excitation signal is applied to the body using a resonator circuit, said resonator circuit including a loop antenna; and Based on the detected change in the electrical characteristics of the resonator circuit, the loop antenna is used to sense the return signal from the body.

[0229] According to another example, a computer program product is provided, the computer program product including code units configured to cause the processor to perform a method according to any of the examples or embodiments outlined above or described below when run on a processor.

[0230] As described above, the system utilizes a processor to perform data processing. Processors can be implemented in various ways, using software and / or hardware, to perform a variety of required functions. Processors typically employ one or more microprocessors that can be programmed using software (e.g., microcode) to perform desired functions. A processor can be implemented as a combination of dedicated hardware performing some functions and one or more programmed microprocessors and associated circuitry performing other functions.

[0231] Examples of circuits that may be employed in various embodiments of this disclosure include, but are not limited to, conventional microprocessors, application-specific integrated circuits (ASICs), and field-programmable gate arrays (FPGAs).

[0232] In various implementations, the processor may be associated with one or more storage media, such as volatile and non-volatile computer memories, such as RAM, PROM, EPROM, and EEPROM. The storage media may be encoded with one or more programs that, when run on one or more processors and / or controllers, perform the desired functions. The various storage media may be fixed within the processor or controller, or may be transferable, allowing one or more programs stored thereon to be loaded into the processor.

[0233] Those skilled in the art, through studying the accompanying drawings, the disclosure, and the claims, will be able to understand and implement variations of the disclosed embodiments when practicing the claimed invention. In the claims, the word "comprising" does not exclude other elements or steps, and the words "a" or "an" do not exclude a plurality.

[0234] A single processor or other unit can implement the functions of several items as described in the claims.

[0235] Although some measures are described in different dependent claims, this does not mean that combinations of these measures cannot be used advantageously.

[0236] Computer programs can be stored / distributed on suitable media, such as optical storage media or solid-state media supplied together with or as part of other hardware, but can also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems.

[0237] If the term “suitable for” is used in the claims or description, it should be noted that the term “suitable for” is intended to be equivalent to the term “configured as”.

[0238] No reference numerals in the claims should be construed as limiting the scope.

Claims

1. A system (8) for processing electromagnetic signals returned from a body in response to an electromagnetic excitation signal applied to the body in an inductive sensing. The system is configured to receive a signal input indicating a sensed return signal, the return signal corresponding to a signal sensed at the loop antenna of the resonator circuit based on a change in the electrical characteristics of the resonator circuit detected when the circuit is driven to generate the excitation signal. in, The system is configured to implement a signal extraction procedure (30), in which the system is configured to: (32) Detect a first input signal from the sensed return signal, the first input signal being based on the frequency of the sensed return signal. (34) Detect a second input signal from the sensed return signal, the second input signal being based on the sensing amplitude of the sensed return signal. The application (36) signal generation procedure includes generating multiple candidate signals, each candidate signal being formed by a different linear combination of the first input signal and the second input signal, and Application (38) is a signal selection procedure for selecting one of the candidate signals, the selection procedure being based on predefined criteria related to one or more signal characteristics of the input signal, the criteria being configured to isolate signals related to specific physiological sources in the body, the selected signal forming an output signal.

2. The system according to claim 1, wherein, The system also includes an inductive sensing device, which comprises: The resonator circuit (10) includes a loop antenna (12). Signal generation unit (14), adapted to excite the loop antenna to generate the electromagnetic excitation signal; and A signal sensing unit (20) is adapted to use the loop antenna to sense the return signal from the body based on the detected change in the electrical characteristics of the resonator circuit.

3. The system according to claim 1 or 2, wherein, The signal generation program is based on the use of independent component analysis.

4. The system according to any one of claims 1-2, wherein, The signal generation procedure forms the plurality of candidate signals based on the use of a set of predefined signal combination ratios.

5. The system according to any one of claims 1-2, wherein, The criteria for the signal selection procedure include one or more of the following: the frequency of the candidate signal, and the number of maximum and minimum values ​​of the signal within a given time window.

6. The system according to any one of claims 1-2, wherein, The signal extraction procedure also includes generating information output indicating specific physiological phenomena based on the selected candidate signals.

7. The system according to any one of claims 1-2, wherein, The signal extraction procedure includes an additional step of applying a bandpass filter to the input signal prior to the signal generation procedure.

8. The system according to any one of claims 1-2, wherein, The signal extraction procedure includes another signal processing step applied directly after the input signal is detected, the signal processing step being configured to suppress motion artifacts in each of the input signals.

9. The system according to claim 8, wherein, The other signal processing step includes: Receives the fundamental frequency input indicating the motion to be suppressed; and A notch filter is applied to the input signal, the notch filter having an adjustable frequency setting, wherein the notch filter is applied to the input signal at one or more times the fundamental frequency.

10. The system according to any one of claims 1-2, wherein, The selection procedure is configured in at least one mode to select candidate signals for the respiratory rate of the identified object.

11. The system according to any one of claims 1-2, wherein, The system is configured in at least one mode to execute the signal extraction procedure at least twice, wherein the signal selection procedure is configured in the first and second runs to select signals associated with different corresponding first and second physiological phenomena.

12. The system according to claim 11, wherein, At least in the second run, a bandpass filter is applied to the sensed input signal prior to the signal generation procedure.

13. The system according to claim 11, wherein, The signal selection procedure is configured in the first run to select a signal related to the subject's respiratory rate, and in the second run to select a signal related to the subject's heart rate.

14. A method in inductive sensing for processing an electromagnetic signal returned from a body in response to an electromagnetic excitation signal applied to the body, the method comprising: Receive a signal input indicating a sensed return signal, the return signal corresponding to a signal sensed at the loop antenna of the resonator circuit based on a change in the electrical characteristics of the resonator circuit detected when the circuit is driven to generate the excitation signal; as well as Implementing the signal extraction procedure (30) includes: (32) Detect a first input signal from the sensed return signal, the first input signal being based on the frequency of the sensed return signal. (34) Detect a second input signal from the sensed return signal, the second input signal being based on the sensing amplitude of the sensed return signal. The application (36) signal generation procedure includes generating multiple candidate signals, each candidate signal being formed by a different linear combination of the first input signal and the second input signal, and Application (38) is a signal selection procedure for selecting one of the candidate signals, the selection procedure being based on predefined criteria related to one or more signal characteristics of the input signal, the criteria being configured to isolate signals related to specific physiological sources in the body, the selected signal forming an output signal.

15. A computer program product including code units configured to cause the processor to perform the method of claim 14 when run on a processor.

Images