[go: up one dir, main page]

WO2024186741A1 - Système et procédé d'identification d'interrogation d'impédance radiofréquence (rfii) - Google Patents

Système et procédé d'identification d'interrogation d'impédance radiofréquence (rfii) Download PDF

Info

Publication number
WO2024186741A1
WO2024186741A1 PCT/US2024/018362 US2024018362W WO2024186741A1 WO 2024186741 A1 WO2024186741 A1 WO 2024186741A1 US 2024018362 W US2024018362 W US 2024018362W WO 2024186741 A1 WO2024186741 A1 WO 2024186741A1
Authority
WO
WIPO (PCT)
Prior art keywords
signals
hemodynamic
rfii
subject
processor
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/US2024/018362
Other languages
English (en)
Inventor
Marc O. GRIOFA
Phil HAMSKI
Robert Rinehart
Dios Aba
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Seemedx Inc
Original Assignee
Seemedx Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Seemedx Inc filed Critical Seemedx Inc
Publication of WO2024186741A1 publication Critical patent/WO2024186741A1/fr
Anticipated expiration legal-status Critical
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/15—Biometric patterns based on physiological signals, e.g. heartbeat, blood flow
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/02—Detecting, measuring or recording for evaluating the cardiovascular system, e.g. pulse, heart rate, blood pressure or blood flow
    • A61B5/024—Measuring pulse rate or heart rate
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
    • A61B5/0507—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves using microwaves or terahertz waves
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/08—Measuring devices for evaluating the respiratory organs
    • A61B5/0816—Measuring devices for examining respiratory frequency
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/117—Identification of persons
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis
    • A61B5/7246—Details of waveform analysis using correlation, e.g. template matching or determination of similarity
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31—User authentication
    • G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints

Definitions

  • the disclosure relates to a system and method for identifying a person based on hemodynamic characteristics and in particular to a radio frequency impedance interrogation (RFII) hemodynamic characteristic system and method for determining an identity of a person.
  • RFIDI radio frequency impedance interrogation
  • biometric based techniques may include fingerprints, voice scans, retinal scans and the like. Some of these techniques are invasive and/or require a user that is awake or able to respond to commands. It is desirable to have an identification system that can measure a biometric and identify a user while the user is unable to respond, unable to follow commands and/or is unconscious.
  • Hemodynamic characteristics of a user may include heart rate (HR) and heart rate variability (HRV), for example. It is known that the monitoring of hemodynamic function may be performed without invasive techniques or skin contact. Potential applications of such technologies include rapid assessment of hemodynamic status, improvement of early medical intervention, and continuous monitoring of subjects without risks associated with invasive techniques. [0005] Hemodynamic characteristics of a user also may be measured/determined using non- invasive techniques. For example, initial investigations have successfully demonstrated the use of a non-invasive radio frequency device for monitoring of heart rate (HR) and heart rate variability (H V).
  • HR heart rate
  • HRV heart rate variability
  • Radio frequency impedance interrogation measures the reflected radio/electrical signal strength from anatomical structures in motion then processes that signal via complex algorithms to generate instantaneous as well as trending measurements of HR and HRV.
  • the RFII device transmits a single frequency RF tone set between 905 to 925 MHz, with no transmitter modulation.
  • the RFII device may have a transducer antenna probe (TAP) that is a specially modified microstrip antenna that is designed to direct the RFII transmitter tone to the blood and tissue of the cardiac mass below the sternum.
  • TAP transducer antenna probe
  • Biological materials are electrically heterogeneous with different tissue types having significantly different complex dielectric constants.
  • blood and blood-filled muscle are tissue entities with the highest dielectric constants, a magnitude higher than the lowest dielectric constant tissues such as bone and fat.
  • the TAP has been specifically designed to match the frequency range for resonance with water-bound hemoglobin molecules.
  • the water-bound protein (hemoglobin) dipole resonance, modulated by the mechanical cardiac activity, is received by the bidirectional TAP.
  • a cardiosynchronous waveform is generated in the time, frequency and voltage domains through the movement of blood from the left ventricle through the aorta, resulting in In-Phase (I) and Quadrature (Q) demodulator channels.
  • FIG. lb illustrates the close correlation of the RFII signal waveform and the known ECG and ICG waveforms.
  • the TAP may be tuned for high return loss, detecting minimal changes in reflected energy from human tissue interfaces, enabling detection of the RF resonance within the thorax.
  • changes in blood volume and cardiac motion will create small deviations in the resonant coupling frequency, thus modulating the RFII transmitter frequency tone.
  • the modulation of the RFII transmitter tone is at a maximum level at or near the resonant coupling frequency since a small deviation in resonant frequency creates a significant phase change.
  • the RFII system enhances resonance of the modulated phase change.
  • ECG signals may be used for subject identification in which short time sequence ECG may be used. These ECG signal based identification systems met with limited success with reported identification rates of 83-99 percent.
  • a RFII device can be used to determine hemodynamic characteristics that may be used for biometric identification of the user and it is to this end that the disclosure is directed.
  • Figure la shows a temporal correlation between an electrocardiogram (ECG) signal and a radio frequency impedance interrogation (RFII) signal for a person;
  • ECG electrocardiogram
  • RFIDI radio frequency impedance interrogation
  • Figure lb shows the correlation of an RFII measurement signal waveform and the ECG and impedance cardiography (ICG) waveforms wherein the channels in Figure lb, top to bottom, are RFII I waveform, ECG, dZ/dt (ICG), RFII second derivative;
  • FIG. 2 illustrates a radio frequency impedance interrogation (RFII) and identification system with a patient interface
  • Figure 3 illustrates an example placement of the patient interface on a human body
  • Figure 4A illustrates a first embodiment of the RFII identification system in which an RFII device is separate from an RFII ID system
  • Figure 4B illustrates a second embodiment of the RFII identification system in which the RFII device is integrated with the RFII ID system;
  • Figure 6 illustrates more details of the RFII main board in Figure 5;
  • Figure 7 illustrates more details of a high isolation full duplexer that connects the RFII device to the patient interface;
  • Figure 8 illustrates an implementation of the patient interface that is a transducer- Antenna-Probe
  • Figure 9 illustrates the average RFII signal Power Spectral Density
  • Figure 10 is the same average power spectral density plot as Figure 9 but shows only the frequencies from 0 to 15 Hz;
  • Figure 11 is a histogram of number of test sessions per subject
  • Figure 12 is an example 5 minute RFII recording for a user
  • Figure 13 shows the differences in output from the frequency tuning step in the RFII recording session in Figure 12;
  • Figure 14 shows an ECG and RFII during a Premature Ventricular Contraction
  • Figure 15 is a plot of instantaneous heart rate from a subject with AF on the left and a normal heart rhythm on the right;
  • Figure 16 illustrates average heartbeat patterns from 4 sessions of subject #3
  • Figure 17 illustrates a method for subject identification and authentication using RFII heart signals
  • Figure 18 illustrates the raw RFII signal and the respiration signal with a top chart showing the raw RFII signal with the effects from respiration clearly visible while a lower chart in which the respiration portion of the spectrum has been subtracted;
  • Figure 19 shows a comparison of extracted heartbeats with respiration (top chart) and heartbeats without respiration compensation (bottom chart);
  • Figure 20 illustrates an impulse response of the low pass filter used for noise filtering
  • Figure 21 illustrates the frequency response of the low-pass filter used for noise filtering
  • Figure 22 illustrated the RFII signal before and after filtering
  • Figure 23 illustrates approximately 200 ms of RFII data with noise filtering;
  • Figure 24 shows a few seconds of RFII data and corresponding ECG R points;
  • Figure 25 show typical Cepstral outputs for heart rate estimation on two different subjects
  • Figure 26 shows a typical spectrogram outputs for heart rate estimation on two different subjects
  • Figure 27 is a histogram of HR estimation errors for the two estimation algorithms
  • Figure 28 is a plot of HR estimation error vs. segmentation error
  • Figure 29 shows a Morlet wavelet on the left and its frequency response on the right
  • Figure 30 shows three RFII signals before and after Morlet Wavelet Filtering
  • Figure 31 shows Extracted RFII heartbeat signatures before and after tapering with the tapered signatures shown in blue and the tapering function shown in black;
  • Figures 32 and 33 show a MACE correlation surface for a positive example and a negative example, respectively;
  • Figure 34 illustrates the k-Nearest Neighbor classifier
  • Figure 35 shows data colored by class (left) and 1 -NN decision boundary colored by class (right);
  • Figure 36 illustrates a Support Vector Machine decision boundary
  • Figure 37 illustrates the construction of the feature covariance matrix for cyclostationarity analysis
  • Figure 38 shows sample covariance matrices from three subjects
  • Figure 39 is a sample gallery and test correlation matrices from 3 subjects
  • Figure 40 is a comparison of heartbeat features before (left) and after (right) time scaling
  • Figure 41 is a comparison of correlation matrices before (left) and after (right) scaling;
  • Figure 42 shows a log-polar original correlation matrix (left) and scaled correlation matrix (right);
  • Figure 43 illustrates example DCT and Inverse DCT matrices
  • Figure 44 illustrates another set of example DCT and Inverse DCT matrices
  • Figure 45 illustrates an example DCT matrices of RFII features for one subject
  • Figure 46 are the first eight principal components estimated using the RFII training data set
  • Figure 47 illustrates an example of an RFII recording one heart beat duration and its reconstruction using 5,10 and 20 principal components
  • Figure 48 illustrates an example of an RFII recording one heart beat duration and its reconstruction
  • Figure 49 illustrates eight independent components estimated using our training data set
  • Figure 50 shows six random bases generated for the random projection method
  • Figure 51 shows an example of an RFII feature and reconstruction after projection onto 400 random vectors
  • Figure 52 illustrates six dictionary elements from the KSVD based tuned dictionary
  • Figure 53 illustrates an example of an RFII recording one heart beat duration and its reconstruction using KSVD based dictionary atoms.
  • the disclosure is particularly applicable to measuring heart characteristics using radio frequency impedance interrogation (RFII) signals and one of verifying the identity of a subject (identification process) or proving the identity of the subject, such as to authorize access systems or privileges (an authentication process).
  • RFIDI radio frequency impedance interrogation
  • the subject/user is a human being, but the system and method could be used on other living beings and be within the scope of the disclosure.
  • the identification or authentication of a subject using the measured heart characteristics may be performed and it is in this context that the disclosure will be described. It will be appreciated, however, that the system and method has greater utility since it may be used while measuring the characteristics of other part of the human body and may be implemented in different ways than the exemplary embodiments set forth below that are within the scope of this disclosure.
  • the disclosure has several different embodiments including one embodiment in which RFII signals are used to determine a heart characteristic that is then used as a biometric modality to identify or authenticate a user.
  • a system and method that produces RF signals and generates RFII signals that may be used to determine a characteristic of a body part of the human, such as the heart of the human are provided in which the characteristic of the body part may be utilized is various different ways.
  • a system and method for identifying or authenticating a user using a body part characteristic are provided in which the signals measuring the characteristics of the body part may be generated in various ways.
  • FIG. 2 illustrates a radio frequency impedance interrogation (RFII) and identification system 20 with a patient interface.
  • the system 20 may have an RFII and ID device 22 that is coupled/connected to a patient interface 24.
  • the patent interface may be placed adjacent to a body part, such as the heart, to measure a characteristic of the body part that may be used to identify/authenticate the human being who has a particular body part with particular, unique characteristics.
  • Figure 3 shows the placement of the patient interface when a heart characteristic is being measured.
  • the RFII and ID device 22 may perform the operations of: 1) generating an RFII signal that is used to measure a characteristic of body part such as the heart (the RFII data collection device); and 2) verifying an identify or authenticating a user identity based on the measured characteristic of the body part based on the RFII signal (the biometric ID device).
  • the RFII data collection device works on the principle of electromagnetic near-field resonant coupling where the coupling mechanism to body tissue is through water bound protein dipole resonance that is known in the art.
  • the RFII data collection device 22 may include a Radio Frequency Impedance Interrogator (RFII), The Transducer-Antenna-Probe (TAP), and the High Isolation Full Duplexer (HIFD) which are each described below.
  • RFIDI Radio Frequency Impedance Interrogator
  • TAP The Transducer-Antenna-Probe
  • HIFD High Isolation Full Duplexer
  • Figure 4A illustrates a first embodiment of the RFII identification system 22 in which an RFII device 42 may be separate from an RFII ID system 44.
  • a computer system may be connected to the RFII device 42 and may have a plurality of lines of instructions/software code that are executed by a processor of the computer system and control the RFII device 42 and performs the identification/verification processes based on the characteristics of the body part, such as the heart.
  • the RFII signals that correspond to the characteristics of the body part may be stored in the device 42 and later communicated to the RFII ID system 44 or may be somewhat constantly communicated to the RFII ID device 44. The communication between the two devices may be secured or unsecured.
  • the embodiment of Figure 4 A may be used in a use case in which the characteristics of the body part, such as the heart, are being measured, similar to having your fingerprints collected or retina scan, etc., are the measured and collected to allow later identification authentication of a user.
  • FIG 4B illustrates a second embodiment of the RFII identification system 22 in which the RFII device 42 is integrated with the RFII ID system 44.
  • the RFII ID system 44 may be implemented as a processor/microcontroller, etc. housed in the device 22 that executes a plurality of lines of computer code/instructions to control the RFII device 42 and perform the processes for identifying/authenticating a user using the characteristics of the body part, such as the heart.
  • This embodiment may be used for use cases in which both the body part characteristic of the user is measured and the identity of the user is verified or authenticated.
  • this integrated system may be used to allow/deny access of the user into a secured area, such as ai airport, secured facility, secured room and the like.
  • FIG. 5 illustrates more details of an embodiment of the RFII device 42.
  • the RFII device 42 and ID device 44 are separated as shown in Figure 4A described above and the RFII device 42 communicates with the external ID device 44 (shown as PC with RFII programming and control application 50) using a well known USB type bus and protocol via a USB connector as shown.
  • the embodiment with the external ID device 44/50 may also have a serial/USB transceiver 52 that converts serial signals to USB and vice versa so that the ID device 44/50 and the RFII device 42 can communicate and the RFII device 42 may be controlled.
  • the RFII device 42 may also have a power supply 54 that may be a +5 volt power supply as shown that may be provided by a rechargeable battery or any other known power source.
  • the power source may also have a different voltage that shown in Figure 5.
  • the power from the power supply 54 may be provided to an power on indicator 56, such as as LED, and to an RF bias tee.
  • the power that passes through the power on indicator 56 may provide power to the serial/USB transceiver 52 and an RFII main board 58 (details of which are shown and discussed below with reference to Figure 6).
  • the patient interface also receives the signals indicative of the characteristic of the body part, such as the heart, and sends those signals (at a sampling rate that may be selected) over a RX in lead of the housing 60 back to the RFII main board 58.
  • the TX out and RX in leads may be coaxial.
  • the RFII signals indicative of the characteristic of the body part, such as the heart, are also sent to a RFII four channel board 62 that generates I and Q unfiltered and filtered signals that may be output to a data acquisition system 64.
  • the data acquisition system 64 may be a commercially available Iworx 24 bit analog to digital converter (ADC) as shown in Figure 5.
  • ADC analog to digital converter
  • the data acquisition system 64 may alternatively be a 10 bit ADC that is on board of the RFII main board 58.
  • the RFII device 42 may also have an adaptive tuning controller 66 that may be a varactor diode for DC bias.
  • FIG. 6 illustrates more details of the RFII main board 58 in Figure 5.
  • the radio frequency (RF) architecture of the RFII main board 58 is similar to a known continuous- wave Doppler radar where the RF transmitting signal and the receiver local oscillator share the same frequency generator and time reference source.
  • part of the mail board 58 is an electromatic compatibility (EMC) shielded area that is delineated by the dotted line in Figure 6.
  • EMC electromatic compatibility
  • the EMC shielded area 600 may be formed in a known manner and may house an oscillator crystal 602, such as a 10 MHz oscillator crystal whose output is input to a frequency synthesizer and VCO 604 that may be a commercially available RF integrated circuit that incorporates a frequency synthesizer and voltage-controlled oscillator (VCO).
  • the frequency synthesizer is programmed to work in the 915 Industrial/Scientific/Medical (ISM) frequency band, at frequency settings from 905 to 925 MHz.
  • the frequency synthesizer-VCO is a true phase-lock -loop (PLL) source, which has its frequency fixed by the crystal oscillator 602 and has dual phase locked outputs with excellent frequency and RF power stability. Both transmit (TX) and local oscillator (LO) signals have fixed RF frequency and amplitude levels.
  • the EMC shielded area 600 may also have a radio frequency (RF) amplifier 606 that receives the output from the frequency synthesizer-VCO 604 that is fed to a RF saw filter 608 and RF low pass (LP) filter 610 and the signal is output on the TX out lead to the patient interface to apply the RFII signals to the user to detect the body part characteristics.
  • the LO signal may be fed to I & Q demodulator 611 via an RF amp 612, RF saw filter 614, RF LP filter 616 and Balun 618.
  • the I & Q demodulator 611 is commercially available and may have RSSI and gain control.
  • the output I & Q demodulated signals are fed into an intermediate frequency (IF) active filter and output buffer 620 which in turn outputs the RFII I & Q signals that are used to determine the characteristics of the body part, such as a heart beat of a user, that can be used to identify/authenticate the user since the heartbeat characteristics of the user are unique.
  • the I & Q demodulator 611 may exchange data with and be controlled by a digital automatic gain control (AGC) circuit and microprocessor. Note that the balun 618, I& Q demodulator 611 and IF active filter 620 are outside of the EMC shielded area 600.
  • the signals received from the user on the RX in lead may be fed into an RF low noise amplifier (LNA) 622, an RF saw filter 624, an RF LP filter 626 and an RF amp 628 and input into the I& Q demodulator 611 that processes the received RFII signals to generate the RFII I & Q signals that may be used to determine or authenticate the identity of the user since the received RFII signals for the user are unique to each user.
  • LNA RF low noise amplifier
  • FIG. 7 illustrates more details of a high isolation full duplexer 700 that is in the housing 60 and that connects the RFII device to the patient interface.
  • the high isolation full duplexer 700 has an connection and port to the patient interface (TAP), receives the TX RFII signals from the main board and outputs the RX in signal to the main board.
  • TX signal is amplified and filtered with high performance Surface Acoustic Wave (SAW) filters and microwave low pass filters in the main board to provide a continuous-wave constant frequency RF test tone with amplitude set to provide a 0 dBm radiating power at the TAP, accounting for RF losses in the TAP, HIFD and miniature RF 50 ohm connecting cables as shown in Figure 7.
  • SAW Surface Acoustic Wave
  • FIG 8 illustrates an implementation of the patient interface that is a transducer- Antenna-Probe (TAP).
  • TAP is a specially modified bidirectional microstrip antenna designed to direct the RFII transmitter tone to the blood and tissue of the cardiac mass below the sternum when measuring the heartbeat characteristics.
  • Biological materials are electrically heterogeneous with different tissue types having significantly different complex dielectric constants.
  • Blood and blood-filled muscle are tissue entities with the highest dielectric constants, a magnitude higher than the lowest dielectric constant tissues such as bone and fat.
  • the TAP has been designed to match the frequency range for resonance with water-bound hemoglobin molecules.
  • the function of the TAP is to separate the TX signal, radiating from the TAP to the body, from the received resonant couple signal (RX), modulated in frequency by motion and blood volume changes.
  • the received resonant (RX) signal is approximately 30dB lower in power than the transmitted signal, so the high isolation duplexer (HIFD) must provide a similar amount of isolation, -30dB or better, for good sensitivity. This is possible with commercial surface mount components, the 3dB quadrature hybrid, and careful design and material selection of the Duplexer RF microstrip board.
  • the TAP as shown in Figure 8 is based on a modified quarterwave microstrip antenna design.
  • a conventional antenna is designed to radiate and received RF power in free space.
  • the TAP is specially modified as a resonator such that the TX signal is efficiently coupled to the human thorax, with enough directionality to confine the energy to the cardiac tissue mass below it as shown in Figure 3. Further, the TAP is capable of and receiving the return RX signal.
  • the TAP kite shape geometry shown in Figure 8 may be a preferred embodiment that was selected to provide the best coupling at the resonant frequency.
  • the TAP works at frequencies only near its resonant band (75 ohms near the resonant frequency) and it can be well matched with a single surface mount miniature RF inductor to 50 ohms.
  • the TAP radiating metal surface is closest to the body, with an effective distance to the skin of the thorax of about 4 mm, including the bottom housing and light clothing.
  • the RF is fed through the hole to the top of the TAP where the RF connector and the TAP RF matching element are located.
  • the TAP cannot work through any metal or other good conductor mass placed below it like a metal pen, jewelry, or soldier’s dog tag since close proximity of a large unexpected additional conductor mass may detune the TAP, but it will work on wet skin.
  • the TAP tuning is dependent on its relative distance to the body, and to a lesser degree the shape of the thorax near the sternum. It also is dependent on position in monitoring cardiac movement and should not be moved during a measurement.
  • the system is programmed to step through frequencies close to the tested TAP resonant frequency to determine the best and most consistent frequency to make a cardiac motion reading after initial placement, signaling the operator when the best signal frequency is acquired.
  • the optimum position for the TAP is at the surface of the lower sternum or breastbone, near the heart.
  • the position of the breastbone also aids to couple the RF energy to the heart due to the scalar dielectric constants of the major biological tissue types as shown in Table 1.
  • the breastbone has a significantly lower dielectric constant (10.0) than the heart (51.0 for muscle and 60.0 for blood).
  • the RF energy can reach the heart with signal disruption than if it had to penetrate solid muscle.
  • lung tissue has a lower dielectric constant and less than muscle, it has more loss and a higher dielectric constant than bone.
  • the distance a transverse-electric-magnetic (TEM) plane wave can penetrate these materials at 2.45 GHz can demonstrate the differential of RF loss between blood, muscle, and fat.
  • a wave can penetrate 19 mm in blood, 17 mm in muscle, and 79 mm in fat [2], These distances would increase somewhat proportionally at the lower frequency of 915 MHz.
  • the RFII receiver is very sensitive to changes in RF phase, and to a lesser extent changes in amplitude, at frequencies between 0.2 and 60 Hz, corresponding to mechanical motions of the human heart.
  • the RFII DCD measures changes in RF impedance as received by the TAP. It is a narrowband device, with a return loss of 10 dB or greater, over about 14 MHz, or a 1.5% lOdB bandwidth.
  • Biological materials are electrically heterogeneous; the first two dispersion mechanisms are caused by the electrical material differences (and perhaps ionic charge flow) located at tissue structures interfaces at the lowest frequency ranges in 1 Hz to 10 KHz to intercellular boundaries, which cause capacitive resonances in the ranges from 10 KHz to 100 MHz. Most bio-impedance measurements are conducted between 1 Hz and 100 MHz, usually in the KHz ranges.
  • the a and b dispersions are heterogeneous and dependent on cell and tissue structure as well as fluid levels and hydration between cells.
  • the high frequency relaxation mechanisms, the d and g dispersions are much different: they are pure dipole resonances at the molecular levels.
  • the pure water resonance, the g dispersion at 20 GHz is best known and observed in fluids where the water molecules, strong electric dipoles, move and rotate in resonance to an applied electromagnetic field at 20 GHz.
  • the d dispersion concerns water-bound proteins resonances, as in animal or human tissues, where the proteins are much larger than water molecules and therefore the resonance frequency is lower, ranging from 300 MHz to 2000 MHz.
  • the fixed dipole resonance is not dependent on tissue or cell structure but instead is very dependent on molecular material content, especially water and protein.
  • the resonance frequency can also be perturbed or changed in time by relative internal motion or volume change.
  • the RFII transmitter tone is modulated, therefore, by the moving tissue and blood flow.
  • the modulation frequencies are very low, starting well below 1.0 Hz for respiration and a clear cardio synchronous signal up to the low end of the audio range.
  • the TAP receives the modulated resonant frequency response signal with the same coupling and directivity of the transmitted single tone frequency.
  • the return RF signal is approximately -30dB lower, or about 1/1000 th of the power of the single tone TX signals.
  • the RX signal is amplified through solid-state, low noise amplifiers, with a gain of about 30 dB, and carefully filtered with high performance SAW filters and microwave low pass filters.
  • the local oscillator (LO) is similarly amplified and filtered, and has an RF BALUN that creates a clean differential signal.
  • the demodulator is an I and Q design, with an “in-phase” and “quadrature,” or 90 degree phase difference, twin outputs for a “sine and cosine” output design.
  • This type of receiver used in RF instrumentation network analyzers, can extract both amplitude and phase information.
  • Radio frequency impedance interrogation refers to the use of a signal near 900 MHz to obtain non-invasive measurements of cardiopulmonary activity or activity of another part of the body.
  • the mechanism whereby the RFII antenna (example of which is shown in Figure 8) identifies internal hemodynamics is based upon the interaction between the 900 MHz signal and the water molecules in the blood plasma and red blood cells. This is in contrast to the typical electrocardiogram (ECG), which measures cardiac electrical activity. It should not be expected that the RFII signal resemble the typical ECG pattern.
  • Figure 9 shows the average power spectral density of the RFII recordings from the supine position of a user.
  • the power spectral density was computed using the known Welch’s method.
  • the 60 Hz noise is greater in power than any information contained in the RFII signal at frequencies higher than about 13 Hz.
  • Figure 10 shows the same RFII power spectral density from 1 to 15 Hz. This is the portion of the spectrum above the 60 Hz noise floor. At the lowest frequencies, two peaks appear corresponding to the average respiration rate and average heart rate for the supine recordings. Heart rate and respiration rate are not unique and change over short intervals of time. The information in this portion of the spectrum (from 0 to 1 Hz) is important but it is not distinguishing of an individual.
  • the frequency band potentially useful for subject identification is roughly from 2 to 15 Hz.
  • the RFII device records both an in-phase and quadrature channel, abbreviated as the I & Q channel respectively.
  • the I & Q channels provide the information necessary to recover the frequency and phase response of the cardiopulmonary system. In the illustrations and plots in this disclosure, it is often easier to depict only one of the two channels as they tend to be similar in appearance and. by convention, when only one channel is shown it is the I channel.
  • the RFII device used has the capability to adjust the transmission frequency of the radio signal. It is constrained to be within about 10 MHz of the resonant frequency of the antenna. Due to variability in manufacturing there are some differences between the resonant frequencies of each antenna. Both of the antennas used in the human subject testing have resonant frequencies of 905.0 Hz. The frequency emitted by the device can be adjusted in increments of 0.1 MHz.
  • the human subject testing conducted comprised of the collection of data from 100 subjects across different stressors to evaluate the changes in the RFII signal that would allow for a more robust biometric. There were 249 tests completed in total. Not all subjects completed all tests, nor were all subjects or tests able to be included in the testing, research and development of each matching algorithm. Data was collected in a supine baseline position and then on a tilt table to induce postural change, a Lower Body Negative Pressure (LBNP) to induce a hemodynamic physiologic change, an exercise induced heat stress to induce a temperature change and a Stroop color test to induce anxiety.
  • LBNP Lower Body Negative Pressure
  • the research database used for final matching process testing and development contains recordings from 78 unique subjects. Most subjects only participated in one test session.
  • the data spans the following protocols: supine, tilt table, lower-body negative pressure (LBNP), exercise, and Stroop color test as shown in Table 2 below.
  • the LBNP protocol is designed to study the effect of hemorrhaging on the biometric signature.
  • the Stroop color test examines the effect of mild stress.
  • the tilt table protocol controls for change in posture and the exercise protocol examines the effect of increased heart rate and core body temperature.
  • Each of the five test protocols contains a baseline supine recording. No subject participated in any of the non-supine protocols more than once. This places a limit on the types of classification experiments that can be performed on the different stressors as will be seen later in this report.
  • nomenclature is given as follows: a recording is a five minute segment of RFII data, a protocol is a specification for a test session (which includes the number of recordings, the posture of the subject during the test, and the environmental stressors to be applied), and a test session is an instance of a test protocol conducted on a particular subject at a particular time, typically containing multiple recordings.
  • FIG 12 shown an exemplary 5 minute RFII recording. Each five minute recording is split into three parts. The first two minutes contain a continuous recording of the RFII signal. For the next ninety seconds, the device sweeps over ten different frequencies in increments of 1 MHz spending nine seconds at each frequency. These frequencies are centered on the natural TAP frequency (NTF) of the device. The frequency sweep portion of the recording is very distinctive, as seen in Figure 13. After completing the frequency sweep, the device returns to the frequency selected during the frequency tuning step for an additional ninety seconds to complete the five minute recording.
  • NTF natural TAP frequency
  • the subject Before each test session the subject underwent a frequency tuning step, during which the device attempts to select the best transmission frequency to obtain a high quality signal.
  • the frequency tuning searches frequencies in 0.1 MHz intervals around the NTF of the RFII device.
  • the device captures a snapshot of the signal at each frequency.
  • each snapshot is ranked based on how close the I & Q channels are to being in- phase and how much of the available range (from 0 to 3 Volts) the signal uses. Frequencies are penalized if they use more than 83% or less than 17% of the available range.
  • the top ranking frequency is selected for the remainder of the test protocol.
  • the frequency tuning step often chooses similar frequencies when it is run on the same person multiple times. However, occasionally it will choose a frequency that differs by a significant margin from what it has previously chosen.
  • the histogram in Figure 13 shows the deviation from the mean frequency selected by frequency tuning. The histogram is relative to the average frequency on a per subject basis. When the frequency tuning selects a frequency that differs from the enrollment frequency, it can degrade the performance of the matching algorithms.
  • Segmentation quality measures the difference in area between the heart rate variability (HRV) curve as calculated by the ECG and the same curve as estimated by the wavelet segmentation algorithm.
  • the signal quality measure looks at the similarity of adjacent heartbeat patterns to each other. In a high quality signal, these signals should be similar.
  • the anthropometric variables used were subject weight, height, body mass index (BMI), approximate body surface area (BSA), heart rate (HR), thoracic circumference and sex. Additionally, three variables relating to the experimental conditions were included. These are device placement (in centimeters below the suprasternal notch), device id number (Boolean indicator), and RFII frequency.
  • Atrial fibrillation presents a more serious challenge to matching. It is somewhat common (observed in 2 of our 78 subjects) and a chronic condition. It is possible for atrial fibrillation to exist without any serious symptoms.
  • AF the normal rhythm of the heart is disrupted by erratic electrical signals in the atria of the heart. This causes the ventricles to contract sporadically producing an irregular rhythm like the one illustrated in Figure 15.
  • the instantaneous heart rate goes from 35 bpm to 100 bpm from one heartbeat to the next then the RFII segmentation may fail.
  • both plots have the same scale on the y- axis and the rapid and extreme variations in heart rate create a difficult segmentation scenario for the RFII.
  • Subject #3 exhibits AF, although not in every session.
  • the average heart patterns for four of Subject #3’s recording sessions are shown in Figure 16.
  • test sessions #29 and #117 the subject exhibited AF, though the same subject demonstrated normal rhythms in sessions #39 or #40.
  • Sessions #39 and #40 were recorded on the same day half an hour apart- -one week after session #29 and two months before session #117. Notice how the pattern in #117 is distinct from the ones in #39 and #40.
  • FIG 17 illustrates a method 900 for subject identification and authentication using RFII heart signals such as cardiopulmonary activity that is unique to each user such that the identity of the user can be determined or authenticated.
  • the cardiopulmonary activity is a different type of biometric signal (like a fingerprint, retina scan, voice print, that is impossible to fake or transfer to another user like is possible with a contact lens having the retina pattern of another person or film that overlays the fingertips of a user to change his/her fingerprints.
  • this method 900 may be performed by the external ID system 44 or the integrated ID system 44 and may be implemented as a plurality of lines of computer code/instructions that are executed on a processor. Microprocessor, microcontroller and the like.
  • the method 900 may receive the RFII signals indicative of the cardiopulmonary activities (one example of the body part characteristics) from the RFII device 42 continuously or periodically to determine or authenticate the user identity on a real time basis as the RFII signals are being collected, but it can also be provided a full set of RFII signals for a user stored and then determine if the RFII signals authenticate the identity of the user at a different time than when the RFII signals are collected.
  • the method may obtain the RFII signals for the cardiopulmonary activities (902) in real time or time delayed wherein the cardiopulmonary activities are unique for each user and can be used to identify or authenticate each user using biometric identification much like retina scans, fingerprints, etc.
  • biometric identification much like retina scans, fingerprints, etc.
  • the cardiopulmonary activities cannot be so easily transferred or faked like the other biometric parameters.
  • the method may pre- process the RFII signals (904) so that the RFII signals may be a better quality for the identification processes.
  • the pre-processing of the RFII signals may include the sub-processes of respiration compensation and noise filtering. Both of these sub-processes may be performed before segmentation and feature extraction.
  • a Discrete Cosine Transform may be used to create a high-pass filter with a cutoff frequency of 0.5 Hz.
  • the top half of Figure 10 shows the raw RFII signal with the effects from respiration clearly visible. The extracted respiration signal is shown in red. On the lower half of the figure the respiration portion of the spectrum has been subtracted. Notice how the filtered RFII signal is now primarily flat, which makes it easier to extract features for subject identification.
  • Figure 18 highlights the importance of the respiration compensation step in feature extraction.
  • the top half of the figure shows the segmented beats from a RFII recording where the respiration compensation was not applied.
  • the bottom half of the figure shows the segmented beats from the same signal with respiration compensation. Notice the tighter grouping of the segmented beats in the bottom plot. The looser grouping in the features extracted without respiration compensation makes the features more likely to be confused with other identities during matching.
  • the RFII signal contains mostly noise in the high- frequency portion of the spectrum.
  • a Parks-McClellan low-pass filter was designed with a cutoff at 27 Hz to remove the high-frequency noise from the signal.
  • the frequency response of the designed filter is shown in Figure 20.
  • Parks-McClellan filters are finite impulse response filters and the particular filter used here as a short duration impulse response of about 1/3 of a second.
  • the filter response to an impulse at the 1.5 second mark is shown in Figure 20.
  • the power of the high frequency noise is much less than the power of the noise in the respiration band so the benefit of filtering it is less extreme. However, it is still good practice to filter out the high frequency noise.
  • the filter may be implemented in software to allow more flexibility in the filter design during the initial research. When implemented in software, as it is now, the noise filtering does take a significant fraction of the processing time during segmentation. This filter could be incorporated as part of the hardware to save on processing time.
  • Figures 22 and 23 show the RFII signal before and after noise filtering.
  • the method 900 may perform alignment and segmentation of the RFII signal (906).
  • a wavelet based filter may be used but these processes may be performed using other types of filters.
  • the alignment and segmentation may be performed in hardware or using a plurality of lines of computer code/instructions executed by a processor.
  • the segmentation and alignment process is important when working with any biometric modality.
  • the first step in face recognition is to align the faces based on the eyes and other facial landmarks. If the faces are not aligned they are essentially incomparable. Similarly, if the cardio-signatures in two RFII signals are not aligned then it will be impossible to perform identification. Segmentation is the process of aligning the RFII signal. Since the cardiac signature is essentially repeated on each cardiac cycle, this is desirable as the fundamental feature used during recognition. The goal of segmentation is then to identify a fiducial point in each cycle of the cardio-synchronous RFII signal. Segmentation occurs in two steps. First the subject heart rate (HR) is estimated.
  • HR subject heart rate
  • the estimated heart rate is used to tune the filter parameters for a Morlet wavelet, which is applied to find the segmentation points.
  • RFII signal pre-processing (respiration compensation and noise removal) take place before segmentation.
  • the segmentation results in RFII signals that can be used for identifying/authenticating a user based on the cardiopulmonary activity in one embodiment.
  • the human subject data collected includes simultaneous ECG recordings.
  • the R points on the ECG signal are easy to locate and can be used to judge the quality of the RFII segmentation. As the RFII segmentation algorithms improve they will get closer to matching the ECG segmentation points. ECG segmentation can then be used as a proxy RFII segmentation in order to present the best-case scenario matching results while research on segmentation methods is still ongoing.
  • the ECG R points in Figure 24 mark each heartbeat in the RFII signal and are a good example of the goal of segmentation.
  • the subject heart rate is estimated by looking at the peak of the Cepstrum of the RFII.
  • the Cepstral is a signal processing technique commonly used in speech processing to isolate the dominant frequencies in a 1 -d signal.
  • the Cepstral is computed as the Fourier transform of the log magnitude spectrum shown in the equation below:
  • the two HR estimation algorithms can be directly compared by referencing them to the ground truth HR obtained via ECG.
  • the histogram plot in Figure 27 compares the relative error of the two algorithms on 255 recordings from the baseline supine protocol. Approximately 97% of the heart rates estimated by the spectrogram method on baseline supine recordings lie within 15% of the ground truth. This test includes a few samples of subjects with abnormal heart rhythms and poor quality recordings for which a true estimate of heart rate is not expected.
  • the subject heart rate is used to tune the parameters for a filter, such as for example a known Morlet wavelet.
  • the Morlet wavelet is a sine wave windowed by a Gaussian and functions as a band pass filter.
  • the wavelet decomposition of a signal is defined as:
  • This function decomposes the signal into shifted and scaled versions of the mother wavelet function i/i.
  • wavelets There are many types of wavelets that can be used for continuous wavelet transforms. Through experimentation, the Morlet wavelet was found to produce the best segmentation results. The Morlet wavelet is defined as
  • FIG. 29 An example Morlet wavelet and its frequency response are shown in Figure 29.
  • the Morlet wavelet is constructed so that the center frequency in the pass band corresponds to the estimated heart rate.
  • the RFII signal is then convolved with the wavelet ideally producing an output resembling a sinusoid, i.e. with only one local maximum per cardiac cycle.
  • the wavelet filtered signal is better behaved than the original RFII signal.
  • the local maxima of the filtered RFII are taken as the segmentation points.
  • An extra step ensures that only local maxima with sufficient separation (at least 0.4 seconds) from adjacent maxima are chosen as segmentation points. This reduces the number of spurious segmentation points detected during periods of low signal to noise ratio. If it is desired to segment signals with heart rates exceeding 150 bpm then it will be necessary to adjust this parameter.
  • the signal is “cut” into segments of 1.2 seconds in length centered on each segmentation point.
  • the length of 1.2 seconds was chosen because it is slightly longer than the average cardiac cycle length. A shorter feature would leave out information by ignoring part of the cardiac cycle. Longer features are impractical due to the effect of heart rate variability causing the features to be out of phase with each other as one moves farther away from the segmentation point and into the next cardiac cycle.
  • a tapering function is applied to the feature in order to increase the relative weight of the portion of the feature nearest to the segmentation point.
  • the tapering window used is Tukey’s tapered cosine window.
  • Figure 31 shows the extracted heartbeat patterns before and after tapering. The un-tapered beats are shown in red and the tapered beats are shown in blue. The tapering function is the black line at the top of the plot.
  • the method may perform feature extraction and matching classification of the features to identify or authenticate a user based on cardiopulmonary activities (908).
  • the extracted heartbeat segments of 1.2 seconds form the base of almost all of the matching algorithms and feature extraction techniques disclosed.
  • the first two minutes of each recording provides more than enough heartbeat features for the matching algorithms. There are diminishing marginal returns as one looks at more and more of the signal since the features are highly correlated with each other.
  • the heart rate estimation step of the segmentation does receive some benefit from having a longer continuous signal to work with. For that reason, most of the feature extraction is performed on the first two minutes of each recording since it is the longest continuous section available.
  • the matching/classification may be performed in hardware or implemented as a plurality of lines of computer code/instructions executed on a processor.
  • the matching/classification process 908 may utilize a variety of different classification processes that may include: a k-nearest neighbor process, a minimum average correlation energy (MACE) filter, a support vector machine process, a cyclostationarity analysis process, a principal component analysis (PCA), an independent component analysis (ICA), a random projection method (RP), K-SVD dictionary learning, kernel correlation filter analysis (KCFA) and/or linear discriminant analysis (LDA).
  • MACE minimum average correlation energy
  • PCA principal component analysis
  • ICA independent component analysis
  • RP random projection method
  • K-SVD dictionary learning K-SVD dictionary learning
  • K-SVD dictionary learning kernel correlation filter analysis
  • LDA linear discriminant analysis
  • a preferred classification process may be the MACE correlation filter although the other processes also may be used.
  • a set of training data may be used to train the filter/classifier/model being used wherein the set of training data may be data with examples of people’s hemodynamic data and the biometric signature that is used to identify the user from the hemodynamic data.
  • the set of training data may be a set of hemodynamic data generated identifiers for a plurality of user’s so that the model/filter/classifier can match the hemodynamic data of a user to an identification signature for the user (that is a unique pattern of hemodynamic data signals that may be used to uniquely identify the user.
  • MACE Minimum Average Correlation Energy
  • An ideal correlation output for a positive detection would be a sharp peak at the location of the object and a flat surface everywhere else, as seen in Figures 32 and 33.
  • the MACE filter tries to achieve this by minimizing the average correlation energy (ACE) of the correlation surface while maintaining the constraint that the center value be a specified value. This is because the training images are assumed to be centered, and so the MACE filter applied to the training images should have a centered peak.
  • ACE average correlation energy
  • correlating the filter with a test sample can also be very fast and will be shift invariant.
  • MACE filters can suffer from too much training data as and will fail to generalize and give good results.
  • MACE filters can be used for both identification and verification (both these terms are defined and used in the experiments section later on in this document).
  • the space in bytes required to store the filter is small, approximately 4 kilobytes per subject.
  • the MACE filter has a closed form solution and is efficient to compute and to apply using frequency domain techniques. Training takes about 0.25 seconds per subject on an Intel 17 processor. More than 90% of the training time is used to apply band pass filters to the RFII signal from the device. These filters were originally implemented in hardware but were moved to software on the prototype device in order to explore all filter options. In making a deployment device, these filters could be re-implemented in hardware for a considerable decrease in processing time. k-Nearest Neighbors Process
  • the k-Nearest Neighbors (kNN) process is a data-intensive method for classification. It assigns a class to test samples based on the average class of the k closest samples in the feature space according to some distance metric. This is usually taken as the l 2 norm or Euclidean distance between two vectors defined as
  • Figure 34 illustrates an example of an unknown point (green circle) being compared to its neighbors. Depending on the parameter, k, the point could be classified as either class 1 (red triangle) or class 2 (blue square).
  • is the offset to the origin. If the data is linearly separable, then two hyperplanes can be placed parallel to the separating hyperplane passing through the closest data points in each class as shown in Figure 36.
  • Cyclostationarity analysis was used previously in gait analysis to identify the person with high accuracy.
  • the heartbeats were segmented from the training RFII signal and 50 of them were selected to estimate the covariance matrix.
  • Each heartbeat is made up of 1200 samples, x x through x 12O o-
  • Each element of the covariance matrix is computed as shown in Figure 37 and the equation below.
  • a covariance matrix is computed from the training recording.
  • the entire matrix is reshaped as a vector in order to compute the distance between the two matrices.
  • the Biometricore team also investigated the use of the correlation matrix created from the same segmented beats as the ones used in estimating the covariance matrix.
  • the correlation matrix is found by taking the matrix of heartbeats X and computing X T X. In other words, each element of the correlation matrix is computed as
  • Figure 40 shows a group of segmented heartbeats from a single subject and an artificially scaled version.
  • Figure 42 shows the shift that occurs in the Log-Polar domain.
  • PCA when applied to a given data set, determines the directions of maximum variance in the data. These directions, the principal components (PCs) are determined by computing the covariance matrix £ followed by its eigen decomposition.
  • the directions need to be found such that, if these data samples are projected onto this direction, the variance in the data is maximized.
  • co the direction vector representing the required projection direction.
  • the variance, on projecting the data onto this direction is given by
  • J (co) [00145] S is the covariance matrix of the data.
  • the PCA needs to maximize J (co) subject to the projection vectors co being unit norm (since they are direction vectors) i.e. subject to
  • 1. This can be done by setting up the problem as a Lagrangian and setting the corresponding derivative to zero.
  • the Lagrangian form of this problem can be written as follows
  • ICA is a popular technique employed in the field of biomedical signal processing, used when there is a need to separate multi-channel biomedical signals into their constituent underlying components. It is a technique to extract a set of underlying sources from a set of observed multivariate data in order to build a generative model for the observed data.
  • the individual components are assumed to be mixed linearly and the mixing model is assumed to be unknown.
  • the observed data are taken to be multiple heart beat samples from a single recording.
  • There is a finite number of overlapping activities (components) which are being measured by the RFII data collection device during each heartbeat duration. Assuming there are K unknown sources that contribute to the RFII measurement during every heart beat cycle, a vector of observations xi at a given time instant can be represented as:
  • the mixing model can be represented as
  • s is a Ax l column vector of unknown source signals at a particular time instant during the beat cycle and 4 is the mixing matrix (model).
  • ICs independent components
  • Random projection is the technique of projecting a set of data samples from a high dimensional space to a randomly chosen low dimensional subspace.
  • n dimensional data sample represented as a column vector
  • y [y 1 ,y 2 , ...,y n j T .
  • D can be generated by choosing its entries from a normal distribution and orthonormalizing the columns using Gram-Schmidt process.
  • the projected vector is given by: n
  • V -? D y (6)
  • K-SVD K-singular value decomposition
  • the method determines the corresponding sparse representation in the dictionary.
  • the sample class is picked by two independent means which are reported below including:
  • d is set as the first column ofU (i.e. corresponding to the largest singular value in A ) and r is set as the product of the first (largest) singular value in A and the first column in V .
  • Steps 2, 3 and 4 are repeated for all d 7 until the function in equation (10) falls below a preset threshold value.
  • the resultant dictionary is used during the matching stage.
  • OMP to solve equation (7), with test samples occupying the data matrix.
  • Figure 52 A few examples of optimized dictionary elements using this approach are shown in Figure 52. Examples of signal reconstruction using this approach are shown in Figure 53.
  • MACE Minimum Average Correlation Energy
  • the column vector u contains pre-specified correlation values, with 1 at the origin for the ‘authentic’ class to which the MACE filter corresponds and 0 for all the other images belonging to the ‘imposter’ classes.
  • system and method herein may be achieved via implementations with disparate or entirely different software, hardware and/or firmware components, beyond that set forth above.
  • components e.g., software, processing components, etc.
  • computer-readable media associated with or embodying the present inventions
  • aspects of the innovations herein may be implemented consistent with numerous general purpose or special purpose computing systems or configurations.
  • exemplary computing systems, environments, and/or configurations may include, but are not limited to: software or other components within or embodied on personal computers, servers or server computing devices such as routing/connectivity components, hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, consumer electronic devices, network PCs, other existing computer platforms, distributed computing environments that include one or more of the above systems or devices, etc.
  • aspects of the system and method may be achieved via or performed by logic and/or logic instructions including program modules, executed in association with such components or circuitry, for example.
  • program modules may include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular instructions herein.
  • the inventions may also be practiced in the context of distributed software, computer, or circuit settings where circuitry is connected via communication buses, circuitry or links. In distributed settings, control/instructions may occur from both local and remote computer storage media including memory storage devices.
  • Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and can accessed by computing component.
  • Communication media may comprise computer readable instructions, data structures, program modules and/or other components. Further, communication media may include wired media such as a wired network or direct-wired connection, however no media of any such type herein includes transitory media. Combinations of the any of the above are also included within the scope of computer readable media.
  • the terms component, module, device, etc. may refer to any type of logical or functional software elements, circuits, blocks and/or processes that may be implemented in a variety of ways.
  • the functions of various circuits and/or blocks can be combined with one another into any other number of modules.
  • Each module may even be implemented as a software program stored on a tangible memory (e.g., random access memory, read only memory, CD-ROM memory, hard disk drive, etc.) to be read by a central processing unit to implement the functions of the innovations herein.
  • the modules can comprise programming instructions transmitted to a general -purpose computer or to processing/graphics hardware via a transmission carrier wave.
  • the modules can be implemented as hardware logic circuitry implementing the functions encompassed by the innovations herein.
  • the modules can be implemented using special purpose instructions (SIMD instructions), field programmable logic arrays or any mix thereof which provides the desired level performance and cost.
  • SIMD instructions special purpose instructions
  • features consistent with the disclosure may be implemented via computer-hardware, software, and/or firmware.
  • the systems and methods disclosed herein may be embodied in various forms including, for example, a data processor, such as a computer that also includes a database, digital electronic circuitry, firmware, software, or in combinations of them.
  • a data processor such as a computer that also includes a database
  • digital electronic circuitry such as a computer
  • firmware such as a firmware
  • software such as a computer that also includes a database
  • digital electronic circuitry such as a computer that also includes a database
  • firmware firmware
  • the above-noted features and other aspects and principles of the innovations herein may be implemented in various environments.
  • Such environments and related applications may be specially constructed for performing the various routines, processes and/or operations according to the invention or they may include a general -purpose computer or computing platform selectively activated or reconfigured by code to provide the necessary functionality.
  • the processes disclosed herein are not inherently related to any particular computer, network, architecture, environment, or other apparatus, and may be implemented by a suitable combination of hardware, software, and/or firmware.
  • various general -purpose machines may be used with programs written in accordance with teachings of the invention, or it may be more convenient to construct a specialized apparatus or system to perform the required methods and techniques.
  • aspects of the method and system described herein, such as the logic may also be implemented as functionality programmed into any of a variety of circuitry, including programmable logic devices (“PLDs”), such as field programmable gate arrays (“FPGAs”), programmable array logic (“PAL”) devices, electrically programmable logic and memory devices and standard cell-based devices, as well as application specific integrated circuits.
  • PLDs programmable logic devices
  • FPGAs field programmable gate arrays
  • PAL programmable array logic
  • electrically programmable logic and memory devices and standard cell-based devices as well as application specific integrated circuits.
  • Some other possibilities for implementing aspects include: memory devices, microcontrollers with memory (such as EEPROM), embedded microprocessors, firmware, software, etc.
  • aspects may be embodied in microprocessors having software-based circuit emulation, discrete logic (sequential and combinatorial), custom devices, fuzzy (neural) logic, quantum devices, and hybrids of any of the above device types.
  • the underlying device technologies may be provided in a variety of component types, e.g., metal-oxide semiconductor field-effect transistor (“MOSFET”) technologies like complementary metal- oxide semiconductor (“CMOS”), bipolar technologies like emitter-coupled logic (“ECL”), polymer technologies (e.g., silicon-conjugated polymer and metal -conjugated polymer-metal structures), mixed analog and digital, and so on.
  • MOSFET metal-oxide semiconductor field-effect transistor
  • CMOS complementary metal- oxide semiconductor
  • ECL emitter-coupled logic
  • polymer technologies e.g., silicon-conjugated polymer and metal -conjugated polymer-metal structures
  • mixed analog and digital and so on.

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Biophysics (AREA)
  • Public Health (AREA)
  • Theoretical Computer Science (AREA)
  • Veterinary Medicine (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Animal Behavior & Ethology (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Physiology (AREA)
  • Cardiology (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Pulmonology (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Software Systems (AREA)
  • Computer Hardware Design (AREA)
  • Radiology & Medical Imaging (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Signal Processing (AREA)
  • Psychiatry (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)

Abstract

La divulgation concerne un système et un procédé permettant d'identifier une personne sur la base de caractéristiques hémodynamiques et en particulier un système et un procédé de caractéristiques hémodynamiques d'interrogation d'impédance radiofréquence (RFII) permettant de déterminer l'identité d'une personne. Le système et le procédé divulgués permettent de générer des activités cardiopulmonaires sur la base d'une interrogation d'impédance radiofréquence et d'utiliser les activités cardiopulmonaires en tant que paramètre biométrique pour déterminer une identité ou pour authentifier l'identité d'un utilisateur.
PCT/US2024/018362 2023-03-03 2024-03-04 Système et procédé d'identification d'interrogation d'impédance radiofréquence (rfii) Pending WO2024186741A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US202363449732P 2023-03-03 2023-03-03
US63/449,732 2023-03-03

Publications (1)

Publication Number Publication Date
WO2024186741A1 true WO2024186741A1 (fr) 2024-09-12

Family

ID=92675480

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2024/018362 Pending WO2024186741A1 (fr) 2023-03-03 2024-03-04 Système et procédé d'identification d'interrogation d'impédance radiofréquence (rfii)

Country Status (1)

Country Link
WO (1) WO2024186741A1 (fr)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090240133A1 (en) * 2006-09-21 2009-09-24 Noninvasive Medical Technologies, Inc. Apparatus and method for non-invasive, in-vivo, thoracic radio interrogation

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090240133A1 (en) * 2006-09-21 2009-09-24 Noninvasive Medical Technologies, Inc. Apparatus and method for non-invasive, in-vivo, thoracic radio interrogation

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
CHANDRASEKHAR BHAGAVATULA ; SHREYAS VENUGOPALAN ; REBECCA BLUE ; ROBERT FRIEDMAN ; MARC O GRIOFA ; MARIOS SAVVIDES ; B.V.K. VIJAYA: "Biometric identification of cardiosynchronous waveforms utilizing person specific continuous and discrete wavelet transform features", ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY (EMBC), 2013 34TH ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE, IEEE, 28 August 2012 (2012-08-28), pages 4311 - 4314, XP032463892, ISSN: 1557-170X, DOI: 10.1109/EMBC.2012.6346920 *
M. O. GRIOFA ; R. S. BLUE ; R. FRIEDMAN ; P. HAMSKI ; M. BHAGAVATULA ; A. JAECH ; SI YING HU ; M. SAVVIDES: "Preliminary feasibility analysis of remote subject identification during hemodynamic monitoring by Radio Frequency Impedance", ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY,EMBC, 2011 ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE, IEEE, 30 August 2011 (2011-08-30), pages 2590 - 2593, XP032319285, ISBN: 978-1-4244-4121-1, DOI: 10.1109/IEMBS.2011.6090715 *
MARC O GRIOFA ; REBECCA BLUE ; ROBERT FRIEDMAN ; KENNETH COHEN ; PHILIP HAMSKI ; ANDREW PAL ; ROBERT RINEHART ; TOM MERRICK: "Radio Frequency Impedance Interrogation monitoring of hemodynamic parameters", BIOMEDICAL SCIENCES AND ENGINEERING CONFERENCE (BSEC), 2011, IEEE, 15 March 2011 (2011-03-15), pages 1 - 4, XP031945032, ISBN: 978-1-61284-411-4, DOI: 10.1109/BSEC.2011.5872326 *
O GRIOFA MARC, REBECCA S. BLUE, ROBERT FRIEDMAN, AARON JAECH, MADHU BHAGVATULA, SIYING D. HU, MARIOS SAVVIDES: "Radio Frequency cardiopulmonary waveform for subject identification", 2011 CONFERENCE RECORD OF THE FORTY FIFTH ASILOMAR CONFERENCE ON SIGNALS, SYSTEMS AND COMPUTERS (ASILOMAR), 6 November 2011 (2011-11-06) - 9 November 2011 (2011-11-09), pages 2152 - 2156, XP093208140 *

Similar Documents

Publication Publication Date Title
Uwaechia et al. A comprehensive survey on ECG signals as new biometric modality for human authentication: Recent advances and future challenges
Paterniani et al. Radar-based monitoring of vital signs: A tutorial overview
US20230267758A1 (en) System and method for detecting movement of an individual using millimeter wave radar
Irvine et al. eigenPulse: Robust human identification from cardiovascular function
Rathore et al. A survey on heart biometrics
Singh et al. Evaluation of electrocardiogram for biometric authentication
Singh et al. Fusion of electrocardiogram with unobtrusive biometrics: An efficient individual authentication system
US10154818B2 (en) Biometric authentication method and apparatus
Abdeldayem et al. ECG-based human authentication using high-level spectro-temporal signal features
Chamatidis et al. Using deep learning neural networks for ECG based authentication
US10380444B2 (en) 3D mask face anti-spoofing with remote photoplethysmography
Tsai et al. Iris recognition using possibilistic fuzzy matching on local features
Pal et al. ECG biometric recognition
Wang et al. Your breath doesn't lie: multi-user authentication by sensing respiration using mmwave radar
Li et al. A novel tensorial scheme for EEG-based person identification
Huang et al. Learning joint and specific patterns: A unified sparse representation for off-the-person ECG biometric recognition
WO2018152711A1 (fr) Système de commande de porte basé sur une authentification électrocardiographique et procédé d'authentification pour celui-ci
Shi et al. Contactless person identification using cardiac radar signals
WO2018032599A1 (fr) Procédé et dispositif d'authentification d'identité destinés à un dispositif intelligent pouvant être porté
Venugopalan et al. Analysis of low-dimensional radio-frequency impedance-based cardio-synchronous waveforms for biometric authentication
Wang et al. Simultaneous authentication of multiple users using a single mmWave radar
Conroy et al. Heart id: Biometric identification using wearable mimo rf heart sensors
Zellmer et al. Highly accurate ECG beat classification based on continuous wavelet transformation and multiple support vector machine classifiers
WO2024186741A1 (fr) Système et procédé d'identification d'interrogation d'impédance radiofréquence (rfii)
Pal et al. Biometric recognition using area under curve analysis of electrocardiogram

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 24767687

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE