+

WO2014153265A1 - Analyse spectrale de flux sanguin turbulent d'artère coronaire - Google Patents

Analyse spectrale de flux sanguin turbulent d'artère coronaire Download PDF

Info

Publication number
WO2014153265A1
WO2014153265A1 PCT/US2014/029833 US2014029833W WO2014153265A1 WO 2014153265 A1 WO2014153265 A1 WO 2014153265A1 US 2014029833 W US2014029833 W US 2014029833W WO 2014153265 A1 WO2014153265 A1 WO 2014153265A1
Authority
WO
WIPO (PCT)
Prior art keywords
vibrational
frequency power
frequency
cardiac data
power spectrum
Prior art date
Application number
PCT/US2014/029833
Other languages
English (en)
Inventor
Norman Lee OWSLEY
Roger Paul NORRIS
Ralph Walter ZAORSKI
Original Assignee
PhonoFlow Medical, LLC
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
Priority claimed from US13/815,961 external-priority patent/US8961427B2/en
Application filed by PhonoFlow Medical, LLC filed Critical PhonoFlow Medical, LLC
Priority to EP14770732.7A priority Critical patent/EP2976016A4/fr
Priority to CA2907400A priority patent/CA2907400A1/fr
Publication of WO2014153265A1 publication Critical patent/WO2014153265A1/fr

Links

Classifications

    • 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
    • A61B5/026—Measuring 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
    • A61B5/02007—Evaluating blood vessel condition, e.g. elasticity, compliance
    • 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
    • A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
    • 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
    • A61B5/7271—Specific aspects of physiological measurement analysis
    • A61B5/7275—Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/08—Clinical applications
    • A61B8/0883—Clinical applications for diagnosis of the heart
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A61B8/5223—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data for extracting a diagnostic or physiological parameter from medical diagnostic data
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B2505/00—Evaluating, monitoring or diagnosing in the context of a particular type of medical care
    • A61B2505/05—Surgical care
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B2562/00—Details of sensors; Constructional details of sensor housings or probes; Accessories for sensors
    • A61B2562/02—Details of sensors specially adapted for in-vivo measurements
    • A61B2562/0204—Acoustic sensors
    • 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
    • A61B5/0803—Recording apparatus specially adapted therefor
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Measuring devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb
    • A61B5/1102—Ballistocardiography
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/103—Measuring devices for testing the shape, pattern, colour, size or movement of the body or parts thereof, for diagnostic purposes
    • A61B5/11—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb
    • A61B5/113—Measuring movement of the entire body or parts thereof, e.g. head or hand tremor or mobility of a limb occurring during breathing
    • 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
    • A61B5/7253—Details of waveform analysis characterised by using transforms
    • A61B5/7257—Details of waveform analysis characterised by using transforms using Fourier transforms
    • 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
    • A61B5/7271—Specific aspects of physiological measurement analysis
    • A61B5/7282—Event detection, e.g. detecting unique waveforms indicative of a medical condition
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/44—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device
    • A61B8/4427—Device being portable or laptop-like

Definitions

  • United States Nort- PiOvisioiia Patent Application titled “Spectrum Analysts Of Coronary Artery Turbulent Blood Flow,” filed on March 1 , 2013, Serial Number 13/815,961 is a continuation-in-part of United States Patent Application Serial No. 12/228,058 filed o « August 9, 2008.
  • the invention relates generally to detecting and processing vibrational cardiac data, and more specifically to apparatuses and methods used to detect vibrational cardiac data related to coronary artery disease.
  • Coronary artery disease is a primary precursor of heart attacks, which is a leading cause of death i n the United States. Coronary artery disease is characterized by a deposition of plaque within the coronary arteries, resulting in a condition referred to as stenosis, in which case blood flow is restricted and the oxygen supply to the heart muscle is decreased. Such a deposition of plaque is also referred to as an occlusion. Coronary artery disease can result in heart attack and subsequent physical injury and possible death. This can present a problem.
  • Turbulent blood flow provides a source of vibrational excitation within the body.
  • the vibrational excitation causes energy to propagate through the body and provides a field that can be measured at the surface of the body.
  • Normal body functions such as breathing and the opening and closing of the heart's valves provide high levels of background noise relati ve to the magnitude of the vibrational energy resulting from excitation at areas of stenosis. Such high levels of background noise can frustrate detection. This can present a problem.
  • the body is made up of structures that have very different physical properties which, are distributed as a function of space throughout the body cavity. Some of these structures are lungs, ribs, organs, blood, arteries, fat, etc. These structures present a non-homogeneous media to the propagation of vibrational energy. Such a non-homogenous media can make it difficult to characterize the media sufficiently to f m: focused listening beams while processing the vibrational energy emitted f om the areas of stenosis during a parametric analysis that assumes a known vibrational wave speed. This can present a problem.
  • Figure A illustrates an apparatus, according to one embodiment of the invention.
  • FIG. 1B illustrates a block diagram of a computer system (data acquisition system) in which embodiments of the invention may be used.
  • Figure 2 illustrates a plot, representational of vibrational cardiac data as a function of time for two heart cycles, according to one embodiment of the invention.
  • FIG. 3 illustrates a method for processing vibrational cardiac data, according to embodiments of the invention.
  • FIG. 1 Figure 4 illustrates several channels of vibrational cardiac data, according to an embodiment of the invention.
  • FIG. 1 (0013)
  • Figure 5 illustrates master replica selection according to an embodiment of the invention.
  • Figure 6 illustrates, in one embodiment,, a correlation scan.
  • Figure 7 illustrates, in one embodiment, assembling multiple heart cycles.
  • Figure 8 illustrates, in one embodiment, diastolic intervals selected for further processing.
  • Figure 9 illustrates a two-dimensional space-time frequency power spectrum (orthogonal vibration mode decomposition of the cross-channel power spectral density matrix "CSDM") of vibrational cardiac data, according to one embodiment of the invention.
  • CSDM cross-channel power spectral density matrix
  • Figure 10 illustrates a two-dimensional space-time frequency power spectrum processed for Equivalent Rectangular Bandwidth, according to one embodiment of the invention.
  • Figure 11 illustrates a phantom constructed to simulate b!ood flow through an area of stenosis, according to one embodiment of the invention.
  • Figure 12 illustrates detection of stenosis in a phantom, according to embodiments of the invention.
  • Figure 13 illustrates an Equivalent Rectangular bandwidth (ERB) display of vibrational energy resulting from fluid flow with occluder present in a phantom (area of stenosis), according to one embodiment of the invention
  • Figure 14 illustrates an Equivalent Rectangular bandwidth (ERB) estimate of vibrational energy resulting from fluid flow without occluder in a phantom (healthy condition without stenosis), according to one embodiment of the invention.
  • ERP Equivalent Rectangular bandwidth
  • Figure 15 illustrates an apparatus according to embodiments of the invention.
  • Figure 16 illustrates a method to obtain vibrational cardiac data according embodiments of the invention.
  • Figure 17 illustrates time series plots of vibrational cardiac data, according to embodiments of the invention.
  • Figure 18 illustrates another set of time series plots of vibrational cardiac data collected from a human whose coronary arteries are in a healthy condition, according to embodiments of the invention.
  • Figure 19 illustrates vibrational frequency power spectra estimates corresponding to a time slot 181 shown in Figure 18, according to embodiments of the invention.
  • Figure 20 illustrates a method for averaging vibrational frequency cardiac data, according to embodiments of the invention.
  • Figure 21 illustrates an overlay of space-time a veraged vibrational frequency power spectra estimates corresponding to data from Figure IS, according to embodiments of the invention.
  • Figure 22 illustrate a set of time series plots of vibrational cardiac data collected from a human with coronary artery disease, according to embodiments of the invention.
  • FIG. 23 illustrates a set of vibrational frequency power spectra estimates corresponding to the time memorize trora Figure 22, according to embodiments of the invention.
  • Figure 24 illustrates an overlay of vibrational frequency power spectra estimates from multiple slots corresponding to the human's data shown in Figure 22, according to embodiments of the invention.
  • Figure 25 illustrates a method for identifying a feature related to coronary artery blood flow turbulence using a single human, according to embodiments of the invention.
  • Figure 26 illustrates a comparison of vibrational cardiac data from muitipie humans, according to embodiments of the invention.
  • Figure 27 iliostraies a method for idcntifymg a feature related, to coronary artery blood flow turbulence using multiple humans, according to embodiments of the in ventioo..
  • FIG. 1A illustrates an apparatus generally at 100, according to one embodiment of the invention.
  • a cross-section 1 12 of a human body contains coronary artery 102 ha ving a flow of blood 1 4 passing therethrough.
  • the flow of blood 1 4 can interact with a coronary artery lesion 10B and cause an excitation of the artery wall by known physical means, which include transition to turbulent flow and the corresponding application of forces normal to the surface of the coronary artery.
  • excitation of the coronary artery wall results in vibrational energy 110 propagating to the surface of the ' human 1 1 1 .
  • the term “sensor” is synonymous with the terms “channel” or “sensor channel,” whereby a separate measurement is contemplated. Additionally, the term “sensor” is synonymous with the terms
  • a first sensor's output (a first channel) and a second sensor's output (a second channel) are each available for analysis and each represents a separate measurement of a field quantity of interest, such as the vibration field in a human's body.
  • a field quantity of interest such as the vibration field in a human's body.
  • An array of sensors 1 16 measures the vibration of the surface 1 1 and collects vibrational cardiac data thereby.
  • the array of sensors ! 16 is made up of a genera! number of N sensors (sensing transducers or transducers), in one
  • the number N equals 14 and the spacing between adjacent transducers is one-quarter inch (0.25"). " Those of skill in the art will recognize that the array of N sensors 1 16 can be configured wi th; a different number of sensors, a different sensor width, and or sensor spacing. The example given herein is provided merely for illustration and does not 1 im.it embodiments of the invention.
  • the cross section 1 1 2 of the human presents a non-homogeneous media through which the vibrational energy 1 10 propagates and contains various structures such as ribs, lungs, organs interfaces, muscles, fat, and skin tissue indicated generally by 1 14.
  • the vibrational energy propagates through the non-homogeneous media and is measured on ( he surface 1 1 1 by the array of N sensors 1 16, in one embodiment, it can be desirable to place the array of sensors 1 1 ( > over a person's heart and above a space between adjacent ribs to facilitate detection of the vibrational energy.
  • each sensor of the array of sensors 1 16 is made from a strip of polyvinylidene fluoride (PVDF) film.
  • PVDF polyvinylidene fluoride
  • each strip of PVDF film measures 0.75 inches long, between attachments to a chassis 122, and 0.1 875 inches wide.
  • Each strip of PVDF film is stretched into a flat plane and is anchored at each end by the chassis 122.
  • a pad is placed to provide an area of contact between the skin surface 111 and the strip of PVDF film.
  • a strip of PVDF film 130 having a first end 132 and a second end 134 (which are attached to the chassis 122) and a pad 136 that makes contact with the skin surface l l ! .
  • the diameter of the pads is 0.1 875 inches and the thickness of the pads is 0.0625 inches.
  • the sensitivity of the PVDF film along its major axis is 22176 V/uait strain for a PVDF film thickness of 0,028 millimeters.
  • the PVDF film generates a voltage in response to strain imparted from the motion of the skin surface 1 1 1.
  • the chassis 1 2 is made out of metal such as aluminum, in other embodiments the chassis 122 is made out of plastic or another material sufficient to provide the necessary anchor points for the strips of PVDF film.
  • Each sensing transducer is in electrical contact: with at least one preamplifier 120 using connection 1 18, It is advantageous to place a preamplifier proximate to its sensing transduce! in order to minimize the addition of electronic noise.
  • Additional amplification stages can be used and in one embodiment the outputs from the preamplifiers 120 are passed to a bank of ampli bombs (not shown), such as those available from ithaco Corporation Model 451.
  • the outputs of the sensing transducers (army 1 16) are carried in a cable bundle 124 and are processed in a data acquisition system .126 that can contain a graphical user interface (GUI).
  • GUI graphical user interface
  • Vibrational energy 1 10 includes shear wave energy propagation with shear wavelengths on the order of several tens of millimeters, e.g. approximately 40 millimeters at 200 cycles per second and approximately 20 millimeters at 500 cycles per second.
  • FIG. 150 illustrates, generally at 150, a block diagram of a computer system (data acquisition system) in which embodiments of the invention may be used.
  • the block diagram is a high-level conceptual representation and may be implemented in a variety of ways and by various architectures.
  • bus system 1 2 interconnects a Central Processing Unit (CPU) 164, Read Only Memory (ROM) 166, Random Access Memory (RAM) 168, storage 160, display 161 , audio 162, . keyboard 164, pointer 166, data acquisition unit (D.AU) 126, and
  • the bus system 1 2 may be for example, one or more of such buses as a system bus, Peripheral Component interconnect (PCI), Advanced Graphics Port (AGP), Small Computer System Interface (SCSI), Institute of Electrical and Electronics Engineers (IEEE) standard number 1394 (FireVVire), Universal Serial Bus (USB), or a dedicated bus designed for a custom application, etc.
  • the CPU 164 may be a single, multiple, or even a distributed computing resource.
  • Storage 160 may be Compact Disc (CD), Digital Versatile Disk (DVD), hard disks (HD), optical disks, tape, flash, memory sticks, video recorders, etc.
  • the computer system ISO can be used to recei ve vibrational cardiac data via 124 from the array 1 16 of vibration sensors ( Figure 1A). Note that depending upon the actual implementation of a computer system, the computer system may include some, ail, more, or a
  • vibrational cardiac data is received at
  • Such data can be transmitted via communications interface 170 for further processing and diagnosis in a remote location, as illustrated in Figure IB at 172, Connection with a network, such as an intranet or the Internet is obtained via 1 2, as is recognized by those of skill in the art, which enables the data processing device 150 to communicate with other data processing de vices in remote locations.
  • a network such as an intranet or the Internet
  • embodiments of the invention can be implemented on a computer system 150 configured as a desktop computer or work station, on for example a WINDOWS* compatible computer running operating systems such as WINDOWS* XP Home or WINDOWS* XP Professional, Linux, etc. as well as computers from APPLE COMPUTER, Inc. running operating systems such as OS X, etc.
  • embodiments of the invention can be configured with devices such as speakers, earphones, video monitors, etc. configured for use with a Bluetooth communication channel.
  • Figure 2 illustrates, generally at 200, a plot of vibrational cardiac data as a function of time for two heart cycles, according to one embodiment of the invention.
  • a representati ve output from, one of the vibration sensors, from array 116 ( Figure 1 A) is illustrated, where a magnitude of the sensor ' s output is plotted on a vertical axis 204 as a function of time 202.
  • a first heart cycle 206 contains a first peak 208 corresponding to the closure of the mitral and tricuspid valves. This first peak is described in the literature as a "lub" sound when heard through a stethoscope.
  • the first heart cycle 206 contains a second peak at 210, which corresponds to the closure of the two semi-lunar, aortic and pulmonary vai ves at the beginning of diastolic period 212.
  • This second peak is described in the literature as a "dub" sound when heard through a stethoscope.
  • the diastolic period 212 follows the second peak 210.
  • a second heart cycle 226 is produced thereby with the same major features found in the first heart, cycle; a first peak at 228, followed by a second peak at 230, and a diastolic interval (Dl) 232. Successive heart cycles (not shown) will continue to occur as the heart continues to beat.
  • Dl diastolic interval
  • Successive heart cycles (not shown) will continue to occur as the heart continues to beat.
  • blood flow is at a maximum in the coronary arteries and unwanted coronary events, such as the first peaks 208, 228 and the second peaks 210, 230 are separated in time and their effect on the diastolic interval is at a minimum.
  • a nominal duration of the entire heart waveform is .from one h undred and twenty (120) to one hundred and eighty ( 1 0) seconds and is made up of six (6) twenty (20) to thirt (30) second segments.
  • a number of heart cycles is approximately equal to ten (10) to fifteen (35) tunes the number of sensor channels in array N. Such a number of heart cycles is needed to adequately resolve the numerically higher eigenvalues as described below in sections of the following discussion. A shorter duration heart waveform (fewer heart cycles) can be collected if the eigen value range is limited accordingly . Those of skill in the art will appreciate that the entire heart waveform can vary in length and that the examples provided herein are given for illustration only and do not limit embodiments of the invention.
  • the number of heart cycles over which a human test subject can comfortably breath-hold will vary between human test subjects and will depend on many factors such as age, physical condition, etc.
  • vibrational cardiac data is collected during breath-hold, the effects of breathing on the measured vibrational cardiac data are minimized.
  • the number of segments can be adjusted to suite the particular test conditions, given the length of time that the human test subject can breath-hold for and the number of sensor channels in the array N.
  • a human starts and stops the acquisition of the vibrational cardiac data to coincide with acquisition during breath-hold periods.
  • FIG. 1 00531
  • the M sensor array described in Figure IA, is used to measure and process vibrational cardiac energy, which is measured at the surface 1 .1 1 during the diastolic intervals, in one embodiment, such measurement and processing of the vibrational cardiac energy is used to determine whether a plaque deposits) (coronary artery lesion(s)) 108 exists in the human due to coronary arter-' disease. In other embodiments, such processing can be used to detect vibrational energy generated within the human in genera! and not necessarily caused by coronary artery disease.
  • Figure 3 illustrates, generally a 300, a method for processing vibrational cardiac data, according to embodiments of the in vention.
  • a method starts at a block 302.
  • a technician seiects a singie high quaiity channel from the arra of N sensing transducers.
  • a high quaiity channel has a high signal ⁇ to-noise ratio, wherein the sigaal-to-noise ratio i expressed as the ratio between the height of a first peak of a heart cycle and the background level during the diastolic interval and the height of a second peak of the heart cycle and background level of the vibrational cardiac data.
  • FIG. 4 illustrates, generally at 400, several channels 402, 404, 406, and 408 of vibrational cardiac data according to an embodiment of the invention.
  • Channel 6 indicated at 404 is selected as the high quaiity channel, with signal -to-noise ratio metric indicated at 410, 10056 j
  • the vibrational cardiac data from the high quality channel is band pass filtered to suppress energy at frequencies that are above and below the frequency content of the first and second peaks of the heart cycle.
  • the band pass filter operation typically passes energy in the band from approximately 5 cycles per second (Hz) to several tens of Hz,
  • envelope detection can be applied to the vibrational cardiac data from the high quality channel.
  • Envelope detection operation is given by:
  • x( ' t) is the high quality channel vibrational cardiac data time series
  • ahs is the absolute value operator
  • eft is the envelope amplitude.
  • one or more segments of heart cycle data can be collected to provide the entire heart waveform as described above.
  • a master replica is selected from each segment,
  • a master replica is selected from the high quality channel, which was specified at the block 304.
  • the master replica is selected by selecting a heart cycle that is highly representative of a majority of heart cycles within the segment of the heart waveform represented by the high quality channel.
  • the master replica is either a portion of or the entire heart cycle so identified.
  • Figure 4 displays vibrational cardiac data, generally at 400, collected front four (4) different transducer channels, i.e., a channel five (5) at 402 compute a channel six (6) at 404, a channel seven (7) at 406 and a channel eight (8) at 408.
  • the vibrational cardiac data collected from channel six (6) at 404 ( Figure 4) will be used for master replica selection and correlation due to favorable signal-to- noise characteristics as indicated at 410.
  • the data from 404 can be band-pass filtered, a described at the block 306 ( Figure 3) and is displayed as 502 in Figure 5.
  • Figure 5 illustrates, generall at 500, master replica selection according to an embodiment of the invention. A noise burst due to breathing is marked at 504 and the same band-pass filtered data is displayed again at 506 where the master replica (MR) window is indicated at 508.
  • MR master replica
  • the master replica is correlated wi th the high quality channel vibrational cardiac data from which it was selected.
  • This cross-correlation procedure produces a correlation waveform that is a function of the time lag between the master replica and the segment waveform extending over the entire length of the segment minus the time length of the master replica.
  • the correlation waveform ' has local maxima when the master replica is temporally well aligned as a function of time lag with a corresponding high signai-to-noise ra tio portion of the segment waveform. These local maxima establish time reference points that are used to identify the diastolic window and to align successive heart cycles in time, i.e., synchronize, for signal analysis.
  • FIG. 6 illustrates, in one embodiment, a correlation scan, generally at 600, that resulted from the vibrational cardiac data shown at 506 in Figure 5.
  • the process begins by analyzing the correlation data 602 to locate local maxima for all values of time (t) for which the correlation coefficient c(t) is:
  • correlation coefficient c(t) is plotted at 602 as a function of time 604.
  • A. threshold is indicated at 60S.
  • the threshold 608 can be defined by an operator with graphical user interface (GUI) or it can be defined by the system.
  • a time difference is obtained between a correlation peak and the peak that came before it in time. If the time difference is less than a threshold, then the maximum peak value is discarded as a possible heart beat cycle starting time. This process discards all candidate heart cycle starting times for heart cycles with a heart rate greater than a specified threshold. For example, a 0.5 second time difference threshold would disallow heart rates above 120 beats per minute (bptn).
  • the local m xima thai are left are used to identify the heart cycles from which the vibrationai cardiac data will be extracted and processed. Generally lower vahres of correlation coefficient can be observed in interval 606 which correspond with the effects of breathing noise.
  • Figure 7 illustrates, in one embodiment, assembling multiple heart cycles.
  • the local maxima that are identified by the analysis described above in conjunction with the block 3 8 are used to define windows in time as the window starting times.
  • the vibrational cardiac data corresponding to these windows in time are over plotted as illustrated; where amplitude is indicated on an axis 702 and time along an axis 704.
  • Envelope amplitude maxima 712 and 7.14 are followed by a diastolic interval 710.
  • Envelope amplitude maxima 71 and 718 are used to help the identification of the diastolic interval; however it is not mandatory to use all four Envelope amplitude ma ima to locate the diastolic interval 710.
  • a single envelope amplitude maxima and knowledge of the human's heart beat rate are sufficient to identify the diastolic interval 710.
  • a start time 706 and a stop time 70S are placed at the ends of the diastolic interval either by a technician or these indicator can be located
  • a power parameter such as average squared amplitude over the duration of the heart cycle is computed for each heart cycle. Then ail of the average squared amplitude levels are averaged to produce a mean squared amplitude level averaged over all heart cycles over plotted. Each heart cycle's average squared amplitude level is compared to a multiple of the mean squared le vel and is discarded if its value exceeds the multiple of the mean squared level. In one embodiment the multiple is equal to 2.0.
  • This heart cycle waveform pruning operation is used to discard those heart cycles that are contaminated by noise which is likely due to breathing and/or intestinal activity.
  • Figure 8 illustrates, in one embodiment; generally at 800, diastolic intervals selected for further processing.
  • vibrational cardiac data from a high quality transducer channel are plotted with amplitude on an axis 802 and time on an axis 804. Vibrational cardiac data are accepted from 21 diastolic intervals.
  • the intervals marked at 806 are contaminated with excessive noise and are rejected.
  • the vibrational cardiac data that are extracted during the identical time window for all N sensors channels, from the diastolic windows, can be processed as a continuous ensemble of data or the diastolic window can be further partitioned into subintervals or slots as described above.
  • the diastolic window 2 ⁇ 2 is divided into four 4 slots 214 and the next diastolic window 232 i partitioned into four (4) slots 234. Adjacent time slots with the slots 214 or 234 can overlap in time.
  • the slots have fixed starting times relative to the respective diastolic interval and are typically separated by less than one tenth of an average heart cycle (for example, 0, 1 seconds for a 60 beat per minute heart cycle).
  • the length of the slot interval, in number of time samples is taken to be the number of points in a discrete Fast Fourier Transform. (FFT) operation which is performed independently within each slot. This procedure effectively strobes the same time slot number (e.g. 1 , 2, 3, 4, etc.) from each heart cycle for FFT spectrum analysis.
  • the temporal length of an FFT window slot is in the range ofO.1.5+0.1 seconds.
  • Figure 9 illustrates a two- dimensional space-time frequency power spectrum (cross-channel power spectral density matrix "CSDM") of vibrational cardiac data, generally at 900, according to one embodiment of the invention.
  • CSDM cross-channel power spectral density matrix
  • spatial frequency number is plotted on an axis 902 and temporal frequency is plotted on an axis 904. Normalized amplitude is indicated by a grey scale color and a reference key is illustrated at 906.
  • the CSDM is either computed for the entire heart cycle, based on averaging ail heart cycles in the entire heart waveform or it can optionally be computed for the a specific slot number in the heart cycle, hi either ease, the CSDM is computed by placing the complex Fourier spectrum (FFT outputs), obtained by processing the transducer channel outputs, into a four-dimensional matrix indexed as x(n, b, k, m): x ⁇ I , h, k, m)
  • n is the vibration transducer number
  • k is the FFT discrete frequency bin number
  • b is the retained heart beat count
  • m is the slot number
  • R(k, ra) ::: — xib, k, m)*s(b, k, ra)'
  • B is the number of heart beat cycles in the averaging ensemble which can span multiple segments of acquired vibrational transducer data in some embodiments.
  • the value of B will depend on the number of separate transducer channels processed for a gi ven measurement. Generally, a lower bound for the value of B is approximately four (4) times the number of transducers, N. A preferred value for B is eight (8) to ten (10) times N. Those of skill in the art will recognize that the goal in selecting the value for B is to reduce the variance in the estimation of the CSDM matrix, therefore the value of B can be set at various numbers and the val ues of eight (8) to ten ( 1 ) are illustrative and not limiting.
  • Figure 10 illustrates a two-dimensional space-time frequency power spectrum (Temporal-Spatial Spectrum (TSS)) processed for Equivalent Rectangular Bandwidth (ERB) constitute generally at 1 00, according to one embodiment of the invention.
  • temporal frequency is plotted on an axis 1004 and eigenvalue number/index (Spatial frequency number (sv» is plotted on an axis 1002.
  • Relative amplitude .1006 of the data is displayed as a modulation of gray scale.
  • an eigenvalue-eigenvector decomposition (EBD) of the CSDM in each slot and for each FFT frequency bin in the range k ⁇ k ⁇ k ⁇ is computed.
  • This decomposition of the CSDM provides estimates of the b!ood flow turbulence induced noise spectrum level and bandwidth.
  • the smallest f eigenvalues are averaged over all FFT frequency values and then these frequency averages are in turn, averaged over the smallest N f values.
  • Nf is typically ten to thirty percent of the total number of transducer channels, N, and the frequency bins over which frequency averaging is performed are within the range above 100.
  • This two dimensional averaged eigenvalue, ⁇ « is termed the TSS noise floor.
  • the TSS noise floor sets a threshold, over whic an accumulation (summation) of eigenvalues is performed. This accumulation of eigenvalues contains an estimate of the blood flow turbulence induced noise energy.
  • the CSDM eigenvalues are plotted along the vertical axis as a functiono of frequency (horizontal axis) .for the sample case described herein.
  • the estimation of the CSDM presented in Figure 9 has been obtained by time averaging the "slot ⁇ interval processed data over five segments with a total of ninety (90) heart cycles.
  • the 3 ia . 5 tf ⁇ and 7 ih harmonies of the 60 cycle power line artifact are evident in the data.
  • these artifacts are nulled, blocked, and extrapolated through, which effectively notches out the FFT frequency bins as illustrated in the image at 950. Similar processing can be performed on the other slots within the diastolic interval.
  • Rectangular spectral Bandwidth, ERB Rectangular spectral Bandwidth, ERB, for spatial eigenvalue p.
  • the estimated number set C [Cup ⁇ *&>, C S ( P) . ⁇ , , ⁇ character ⁇ , for p - 1 , 2, ... , ⁇ Nf can provide a diagnostic tooi for the detection of arterial blood flow turbulence and thereby the causative pathology.
  • a simulation of such detection was performed on a phantom and is described below in conjunction with Figure II through Figure 14.
  • the process begins by pre-sraoothing the estimated frequency
  • the eigenvalues as a function of frequency, k, that exceed the threshold are given b the averaged value above tie threshold as ⁇ - ⁇ a d those that do not exceed the threshold are counted by C ⁇ and expressed by the averaged value below the threshold as ⁇ ⁇ ) .
  • FIG. 11 illustrates a phantom, generally at 1 100, constructed to simulate blood flow through art area of stenosis, according to one embodiment of the invention.
  • a phantom simulator of the human thorax 1 102 in the vicinity of the chest wail was formed from silicone gel 1 106 with a predcterrained stiffness comparable to human tissue.
  • the phantom contained high- stiffness human rib surrogates 1 10$ made of shaped plastic.
  • Blood flow was simulated within latex tube 1 104 having cross-sectional characteristics typical of the left anterior descending (LAD) human coronary artery with a 3.0 millimeters (mm) inner diameter. Fluid flow occluders of different topologies were inserted into the latex tubing and the blood was simulated in viscosity with mixtures of 25 to 50 % glycol and distilled water. All dimensions were known and the latex tube was embedded at a depth of 40 mm from the contiguous surface of the phantom.
  • the long axis of the vibration sensor linear array (not shown) was placed parallel to the rib surrogates 1108 and directly over the linear space between a pair of the rib surrogates 1 108.
  • Such placement simulates placement on a human and established a direct vibration wave path from the induced turbulent flow site to the v ibration sensors located on the contiguou s surface of the phantom.
  • Figure 12 illustrates detection of stenosis occlusion i a phan tom, generally at 1 200, according to embodiments of the invention.
  • the measured power spectrum level of vibration transducers is plotted on a vertical axis 1204.
  • Spectrum le vel represents an integration in frequency of the energy in a given eigenvalue.
  • Flow speed is plotted on an axis 1202.
  • a series of experiments were conducted with different flow speeds with and without an occluder present in the latex tube 1 1 4 ( Figure 1 1 ) to simulate an area of stenosis in a human.
  • Eigenvalue svl4 is shown at .1206 with occluder and at 1 208 without occluder.
  • Eigenvalue sv0.1 is plotted at 1216, as a function of flow speed, with occluder in to simulate an area of stenosis.
  • Eigen value svOl. is plotted at 12.18 without occluder to simulate the healthy state, free of stenosis.
  • Figure 13 illustrates an Equivalent Rectangular ' bandwidth (ER ' B) display of vibrational energy resulting from fluid flow with occluder present (area of stenosis), generally at 300, according to one embodiment of the invention.
  • temporal frequency is plotted on an axis 1302 and eigenvalue number/inde is piotted on an axis 1304.
  • Relative amplitude 1308 of the data 1306 is displayed as a modulation of gray scale.
  • Data 1306 represents an Equi valent Rectangular Bandwidth (ER ' B) estimate for the 35 cm/see flow rate with an occluder present.
  • Figure 14 illustrates an Equivalent Rectangular bandwidth (ERB) estimate of vibrational energy resulting from fluid flow without occluder (healthy condition without stenosis), generally at 1400, according to one embodiment of die invention.
  • ERP Equivalent Rectangular bandwidth
  • the same format is used to present the data from the 14 channel array, temporal frequency is plotted on an axis 1402 and eigenvalue number/index is plotted on an axis 1404.
  • Relative amplitude 1408 of the data 1406 is displayed as a modulation of gray scale.
  • the flow rate of the simulated blood flow was 72 era/see. Such a rate is higher than what typically exists during normal Mood flow in a healthy human. This high rate (72 cm/sec) was selected for purposes of comparison in order to present a worst case detection scenario for the methods described herein.
  • Figure 13 shows that three spatial frequency eigenvalue modes 1310 are excited with sufficient energy to exceed the 3 db E B threshold for the 35 era see flow velocity.
  • Figure 14 indicates that only two spatial modes 1410 exceed the ERB threshold level at notably small value of the estimated ERB. The discrimination provided by these two extreme cases illustrates the presence of detected turbulent flow induced noise,
  • Figure 15 illustrates, generally at 1500, an apparatus according to embodiments of the invention.
  • Figure 15 a cross-sectional view of a human is presented, in Figure 15, the cross section contains a coronary artery 1502. A direction of blood flow is indicated at 1504.
  • An occlusion irt the artery 1506 participates in causing downstream blood flow turbulence 1508.
  • blood flow turbulence produces a state of vibration 1510 in the human's body which propagate energy in the form of elastic waves 1510.
  • the eiasiic waves produce vibration of the surface of the human and are measured ncm-invasively by one or more sensors 15.16.
  • the elastic wave energy 1510 propagates through the human's body it interacts with internal structures such as lung 1514 and ribs 1512. in various embodiments, it can be desirable to measure the vibration of the surface of the human in-between such ribs 1512, by placing a vibration transducers) in the intercostal space (between ribs) as shown by the placement of sensor 1 6.
  • the surface may provide a curved surface, such as is indicated by 1522.
  • the sensor or array of sensors is placed in such manner as to conform to the geometry of the surface 1522.
  • a variety of types of transducers can be used to measure vibration of a surface of a human.
  • a film sensor such as described above in conjunction with Figure 1 A can be used as well as other sensors, such as for example a laser based sensor.
  • Sensors that respond to displacement, velocity, and acceleration can be used to measure the vibrational energy that manifests on the surface of the human.
  • proximity sensors can be used.
  • Sensors as of now yet unknown can be used to measure the vibration of the surface of the human.
  • Embodiments of the invention ate not limited by the choice of sensor used to non- invasively measure the vibrational energy manifest on the surface of the human. Those of skill in the art will note that if a laser based sensor is used, physical contact between the sensor and the surface of the human is not necessary as a laser can be used to measure the surface without making phy ical contact apart from the laser beam interacting with the surface.
  • Spectrum analyzer 1520 is in various embodiments a single or muhi-ehanneS spectrum analyzer that performs a transformation f om: time to frequency.
  • the resulting transformation has been referred to above in conjunction with Figure 3 as a complex Fourier spectrum of the vibrational cardiac data and the ensemble average of the amplitude squared spectrum is referred to below as a vibrational frequency power spectrum estimate or simply "spectral estimate” or spectral estimates” when plural spectra are referred to.
  • spectral estimate or spectral estimates
  • the vibrational frequency power spectrum estimate is presented on a display 1524 for view by a doctor or technician or the vibrational frequency power spectrum estimate can be analyzed by an automated system to extract features related to a condition of health of a coronary artery.
  • an automated system is configured in various embodiments as described above in conjunction with Figure IB.
  • the vibrational frequency power spectrum estimate is used to quantify features and their characteristics which are used to support a diagnosis of a state of health of a human.
  • a list of fea tures and their characteristics includes, but is not limited to blood flow turbulence energy from: (a) ventricular refilling turbulence characterized by a low frequency (LF) plateau with feature characteristics of level, width, and roll-off rate (slope) from the LF plateau; f b) arterial blood flow turbulence induced by diffuse occlusions characterized by a characteristic that captures the broadband spectrum level increase in the high frequency (HF) band above a 200 to 400 Hz roll-off band relative to the LF plateau level; (e) relatively localized occlusions described as having a band of surface curvature radials that induce spectral energy swaths characterized to first order by the characteristics of spectrum level, center frequency, and bandwidth; (d) localized occlusions with a dominant
  • Figure 16 illustrates, generally at 1600, a method to assess a condition of health of a coronary artery using vibrational cardiac data, according; to
  • a method starts at a block 1602, At a block 1604 vibrational cardiac data is acquired non-invas.fve.ly from a surface of a human. At a block 1 06 unwanted coronary events are separated from the vibrational cardiac data. Separating unwanted coronary events such as heart valve vibrations are described above in conjunction with Figure 3. At a block 1608 a transient event is extracted from the vibrational cardiac data. The transient event is associated with blood flow turbulence and is used to assess a condition of health of a coronary artery. At a block .1610 the transient event is evaluated for an indication of energy due to coronary artery blood flow turbulence and the corresponding state of health of the coronary artery.
  • a feature is identified in the vibrational frequency power spectrum estimate that is associated wit blood flow turbulence. Relating the ieature(s), or characteristic ⁇ ) thereof in the vibratioiiai frequency power spectrum estimate to state of health in the coronary arteries of a human is described below in conjunction with the figures that .follow.
  • the method stops.
  • FIG. 17 illustrates, generally at 1700, time series plots of synchronized vibrational cardiac data, according to embodiments of the invention.
  • time is plotted on a horizontal axis 1702 and amplitude is plotted on a vertical axis 1704.
  • vibrational cardiac data was collected non-mvastvely from a surface of human with a transducer.
  • Displayed in Figure 17 is a stack of 67 synchronized heart-cycle time series waveforms.
  • the stack of 67 heart-cycle time series waveforms is from one channel of a 14 channel array, hi order to maximize signal-to-noise ratio (SNR) in a given measurement, heart cycles can be pruned, thereby discarding outliers that arc contaminated with various sources of noise.
  • SNR signal-to-noise ratio
  • various extraneous stimuli m the environment can manifest themselves as noise in vibrational cardiac data. These stimuli include, but are not limited to, stomach gurgling, etc.
  • the inter- waveform correlation coefficient with a master replica exceeds 0.87 for all of the 67 waveform pairs in Figure 17.
  • FIG. 100102 Similar to the representation ofSieart-cycie waveforms shown above in Figure 2 or Figure 7, two diastolic intervals are shown in Figure 17.
  • a first diastolic interval is observed at 1706.
  • the first diastolic interval 1706 is followed by a systolic interval 1708 and then a second diastolic interval 1710.
  • a first heart sound is indicated at 1732 and a second heart sound ts indicated at 1714.
  • Heart sounds 1712 and 1 14 are some of the unwanted coronary events that are separated from the vibrational cardiac data. During diastole, blood flow in the coronary arteries is at a maximum and the unwanted heart sound vibrations are at a minimum.
  • the vibrational cardiac data that occurs during a diastolic interval are processed to assess a condition of health of the coronary arteries.
  • Blood flow through the coronary arteries is at a maximum at the onset of diastole and then decreases as a function of time through diastole.
  • corresponding state of health of a human's coronary artery can be obtained from an analysis of the blood flow turbulence energy that occurs during diastole.
  • a diastolic interval such as 1706
  • a diastolic interval is divided into a plurality of time slots, such as the example of four time slots shown in Figur 1? as indicated by ⁇ 716, 1.718 tone 1720, and 1722.
  • Transforming the time series waveforms that occur in these time siots to the frequency domain produces vibrational frequency power spectra estimates which are compared in order to obtain information about both the transient and stationary behavior of the blood flow turbulence,
  • a time siot is an interval of time with a known start time and duration time reiative to an established time mark.
  • a time slot is also referred to herein as a sub-interval or a portion of a diastolic interval Note also that a time slot can be configured to be equivalent in duration to a diastolic interval.
  • the time mark can be selected either manualiy or automatically by the system, in one embodiment, a preierred location for a start time is very early diastole durin the rapid early ventricular filling phase, when coronary blood flow is at a maximum rate.
  • a typical time slot length can be in the range 0,125 to 0.1825 seconds in duration when four (4) rime slots are used to process the diastolic window with 50% overlap between time slots.
  • Other amounts of time slot overlap can be used and in some embodiments time slots can be configured without overlap.
  • the example of four (4) time siots with a 50% overlap is provided merely for illustration and does not present any limitation to embodiments of the invention.
  • Figure 18 illustrates, generally at 1800, another set of time series plots of vibrational cardiac data collected from a human whose coronary arteries are in a healthy condition, according to embodiments of the invention.
  • a stack of 1 9 synchronized heart-cycle waveforms are displayed in the figure with time plotted on an axis 1 02 and amplitude plotted on an axis 1804.
  • the correlation coefficient computed between a master replica and all other heart cycles exceeded 0.89 for pairs of heart cycles in the stack of Figure 18.
  • the human corresponding to the data shown in Figure 18 is a forty three (43) year old male who is symptom free with respect to cardiovascular disease.
  • the individual exercises regularly and has a total cholesterol value of 170.
  • the vibrational cardiac data from this individual is used below to provide information on blood flow turbulence from a healthy state of coronary artery health.
  • the first heart sound interval is indicated at 1806,
  • the first heart sound interval 180 includes a closure snap 1812 of a mitral valve and a closure snap of a tricuspid valve at 1814.
  • the second heart sound interval 1808 includes an aortic valve closure 1818 and i 820.
  • 1 820 is either a pulmonary valve closure and/or an early ventricular refilling turbulence transient.
  • the diastolic interval is the region of interest.
  • the aforementioned heart sounds constitute unwanted coronary events and are eliminated from the processing by placement of the time slots.
  • the time intervals that are used for estimation of the vibrational frequency- power spectra are pre first heart sound time slot 1 816, diastolic interval time slot 1 (1822), diastolic interval time slot 2 ( 1824), diastolic interval time slot 3 (1826), diastolic interval time slot 4 (1828).
  • a heart sound referred to as the fourth heart sound is indicated at 1830, which is caused by diastole refilling turbulence transicut.
  • fOO 107J ' he stack of one hundred and nine ( 109) vibrational cardiac data measurements from Figure 18 are transformed to the frequency domain using techniques such as the Fast Fourier Transform (FFT). Other techniques can be used to transform to the frequency domain, Ac FFT is used for example and does not. limit embodiments of the invention.
  • FFT Fast Fourier Transform
  • Figure 19 illustrates, generally at 1 00, vibrational ⁇ frequency power spectra estimates corresponding to a time slot 1816 shown in Figure 18, according to embodiments of the invention.
  • the one hundred and nine ( 109) time slot 1 sub-intervals in the synchronized heart cycles are Fourier transformed and then the amplitude squared of the complex Fourier transform are ensemble averaged for a given channel, in one embodiment, the procedure is repeated for each channel when an array of transducers is used to obtain the vibrational cardiac data.
  • the averaged vibrational frequency power spectrum estimate for each channel is indicated for a channel number at 1 02 and is plotted as a function of frequency on an axis 1904.
  • the corresponding normalized amplitude is indicated by gre scale with amplitude reference given at 1906 in decibels.
  • the lower plot in Figure 19 displays the vibrational fre uency power spectrum estimate obtained by a veraging a subset of channels from .1 02.
  • channels one 1 ⁇ through thirteen (13) from the fourteen (14) channel array are averaged together.
  • Frequency is indicated on a horizontal axis at 1908.
  • Spectrum level is indicated on a vertical axis at 1910.
  • the dominant feature of the spectral estima tes is a low frequency plateau 1920.
  • the low frequency plateau 1920 includes a relatively flat low frequency region of the spectrum between, approximately eighty (80) and. two hundred (200) cycles per second (Hz).
  • valve vibrational energy propagates by means of elastic waves in the walls of the heart chamber, if (here is other energy that is time coincident with the third and fourth heart sounds, e.g., 1820 and 1.830 (Figure 18), then the corresponding spectrum is masked fay 1 820 and 1 30 ( Figure 18). Power line artifacts of 60 Hz are indicated at 1918.
  • FIG. 20 illustrates, generally at 2000, a method for averaging vibrational frequency cardiac data, according to embodiments of the invention.
  • a method starts at a block 2002.
  • a block 2004, in one embodiment as described above in conjunction with Figure 19 an ensemble of vibrational frequency power spectra estimates, corresponding to time slots in tie synchronized heart cycles are averaged.
  • the resulting plurality of time a eraged vibrational frequency power spectra estimates ate averaged together, la one embodiment, an example of such averaging in rime and transducer channel position space is shown in Figure 19 at 19.10.
  • the process stops at a block 200 .
  • averaging vibrational frequency power spectra estimates can be performed on all or a subset of channels.
  • Figure 21 illustrates, generally at 21 0, an overlay of space-time averaged vibrational frequency power spectra estimates corresponding to data from: Figure 18, according to embodiments of the invention.
  • the space-time averaged vibrational frequency power spectrum estimates 2118 are plotted with frequency on an axis 21 2 and spectrum level on an axis 2104.
  • a twenty five (25) dB roll-off indicated at 2106, occurs in the 200-400 Hz band indicated at 21 1.0 for the averaged power spectrum estimate for time slot 2, indicated at 2124.
  • a 28 dB roll-off indicated at 2108 occurs in the 200-400 Hz band 21 10 for the averaged power spectrum estimate for time slot .1 , indicated at 2122,
  • Reference point 2112 is the 3 dB down point from the low frequency plateau.
  • Reference point 21 12 is used to measure the roll-offs described directl above. Artifacts of 60 Hz power iine harmonics are evident at 21 16. Those of skill in the art will recognize that such narrow band 60 cycle artifacts are tolerable as Song as they do not bias the estimated power spectral density (PSD) level outside of their narrow band.
  • PSD estimated power spectral density
  • the spectrum for time slot 1 2122 could contain, energy from the pulmonary valve closure .1 20 ( Figure 18) and the earl ventricle refilling 1830 (Figure 18).
  • the pulmonary valve snap will generate broadband energy above 200 Hz which appears to be at a very low level because this band has the lowest spectrum level for all four time slots 21 18.
  • the ventricle refilling will produce energ below 200 Hz because it results from the flow over the surface of the atrioventricular valves which have relatively large surfaces of order centimeters with low curvatures relative to the inner dimensions of a coronary artery, which has dimensions of order millimeters with proportional curvatures,
  • Time slot 2 captures the trailing edge of ventricle refilling (S3), the leading edge of S4 and a uiet area which permits measurement of energy due to blood flow turbulence in the left coronary artery.
  • a moderate strength spread spectrum energy swath is indicated at 21 14. This swath has a center frequency of 350 Hz, a bandwidth of approximately 60 Hz, and a signal-to-noise ratio (SN ) of approximately 8-10 dB. This
  • Time slot 2 shown at 2124 reckon also indicates a low level of spectrum ripple.
  • the ripple has a period of approximately 30 to 40 Hz and a peak-to- alley amplitude differential of 2 to 3 dB as indicated at 2126.
  • Tin ' s effect is consistent with an interference pattern produced by energy propagating from a vibration source to a vi bration transducer (measurement location) along more than a single path.
  • Phase coherent energy arrivals on different paths can periodically suppress or support each other and a frequency spectrum ripple period of 30 to 40 Hz is consistent with elastic wave propagation speeds in tissue with multiple path length differences on the order of centimeters.
  • Figure 22 illustrates, generally at 2200, a set of time series plots of vibrational cardiac data collected from a clinically diagnosed individual with a history of coronary artery disease, according to embodiments of the invention.
  • the sub ject of the measurements presented in Figure 22 is a male in his late sixties. Six months prior to the measurement, this man was diagnosed with total blockage of Left Anterior Descending (LA D) coronary artery and underwent a stent procedure that restored full L AD flow.
  • LA D Left Anterior Descending
  • Figure 22 are one hundred and four (104) synchronized heart cycle waveforms of vibrational cardiac data. These data have been collected as previously described and, in this embodiment, an array of fourteen (14) transducers was used to collect die vibrational cardiac data.
  • the synchronized heart cycle waveforms exhibit a master replica pair wise correlation coefficient greater than 0.91.
  • time is plotted on an axis 2202 and amplitude is plotted on an axis 2204.
  • a systolic interval is indicated at 2206 and a diastolic interval is indicated at 2208.
  • the diastolic interval 2208 has been partitioned into four (4) overlapping time slots 2210, 2212, 214, and 2216.
  • Transformations from time to frequency are • performed on each heart cycle waveform within each time slot. This process is repeated for each channel of the array of fourteen transducers. [0 119] In one embodiment, for each channel, an ensemble of power spectrum estimates are then averaged for the same time slot in each heart cycle. The results of the averaged rime slot power spectrum estimates (rime to frequency transformations) are shown ia Figure 23 on a channcl-by-cha «nel basis with the time ensemble average at the top and as a channel average across the array at the bottom:,
  • Figure 23 illustrates, generally at 2300,, a set of vibrational frequency power spectra estimates correspondin to the time series from Figure 22, according to embodiments of the in vention.
  • each of the averaged spectral estimates for each channel are displayed at 2300 with frequency on an axis at 2302, channel number on an axis 2304 and normalized spectrum level is indicated at 2305 as gray scale.
  • estimates from 2300 are averaged together and plotted in 2350 with frequency on an axis at 2352 and spectrum level on an axis 2354, Averaging as was done in both time (ensemble of heart cycles) and space (across channels) improves a signal-to-noise ratio of the measurement. Such a veraging is not always required and in some embodiments with sufficient stgnai-to-noise ration (SNR), -meaningful information is obtained from a single time to frequency transformation of a single heart cycle from one transducer.
  • SNR stgnai-to-noise ration
  • the processed vibrational frequency power spectrum estimate plotted in Figure 23 at 2350 corresponds to the first time slot 2210 (Figure 22) in the diastolic interval 2208 ( Figure 22).
  • Features are present in 2350 that are used to identity coronary artery blood flow turbulence .
  • One feature is a low frequency plateau, which exists between approximately 80 and 150 Hz and is indicated at 2356, This low frequency plateau is associated with valve flow as previously discussed.
  • Another feature i a low frequency roil -off from the plateau which occurs between 80 and 150 Hz.
  • the ro!i-off of the amplitude is 18 dB as indicated at 2358 over the 200 to 400 Hz frequency octave.
  • Another feature is a strong spread spectrum energy swath which occurs at 2360. This strong swath is characterized by a center frequency of 230 Hz, a bandwidth of 80 Hz, and a signal-to-noise ratio of 18 dB. Yet another feature is a weak spread spectrum energy swath which occurs at 2362. This weak swath, is characterized by a center frequency of 365 Hz, a bandwidth of 55 Hz, and a signal-to-noise ratio of 7 dB. Another feature is a medium spread spectrum energy swath which occurs at 2364.
  • This medium swath is characterized by a center frequency of 570 Hz, a bandwidth of 1 0 Hz, and a signal-to-aoisc ratio of 8 dB.
  • Another feature is a frequency band-limited whistle at 2366, Whistle 2366 is characterized by a center frequency of 805 H arid amplitude of 8 dB.
  • Another feature is a difference in amplitude between the low frequency plateau 2356 and a high frequency broad band level. This difference is approximately 30 dB as indicated at 2368.
  • Several power line artifacts of 60 Hz are evident at 2370. These power line artifacts are not considered to be features.
  • Figure 24 illustrates, generally at 2400, an overlay of vibrational frequency power spectra estimates from -multiple slots corresponding to the human's data shown in Figure 22, according to embodiments of the invention.
  • time and channel averaged vibrational frequency power spectrum estimates for each time slot (22.1 , 2212, 2214, and 2216 from Figure 22) are plotted o a graph with frequency on an axis 2402 and spectrum level on an axis 2404,
  • Time slot I , time slot 2, and time slot 3 contain features, the types of which were described above, which are associated with coronary artery blood flow turbulence and a state of health of a coronary artery.
  • Time slot 4 (22.16 in Figure 22) contains valve snap energ and has been placed to capture part of the valve snap to illustrate die fact that the first three time slots (2210, 2212, and 2214 from Figure 22) are measuring blood flow turbulence.
  • a feature in time slot 1 is a strong spread spectrum swath 2406.
  • This strong spread spectrum swath 2406 has a center frequency of approximately 230 Hz and a bandwidth of approx imately 90 Hz 2407.
  • time slot 2 (2212) the strong spread spectrum swath 2406 (from time slot I 2210) has transformed into a frequency band limited whistle at 2408 with a center frequency of 200 Hz at 2424.
  • Another example of a feature changing between time slots is medium spread spectrum energy swath 2412 in time slot 2 (2212) transforming into a frequency band limited whistle 2414 in time slot 3 (2214).
  • the estimated center frequency of the swath 2412 and the whistle 2414 is 390 Hz as indicated at 2422.
  • FIG. 25 illustrates, generally at 2500, a method for identifying a feature related to coronary artery blood flow turbulence using a single human, according to embodiments of the .invention.
  • a process starts at a block 2502
  • a di astolic interval of a heart cycle is partitioned into at least two time slots.
  • a time to frequency transformation i performed on vibrational cardiac data collected from the time slots created in the block 2504.
  • monitoring paradigm for heart disease includes periodic
  • FIG. 26 illustrates, generally at 2600, a comparison of vibrational cardiac data from multiple humans, according to embodiments of the in vention.
  • frequency is plotted on an axis at 2602 and spectrum level is plotted on an axis at 2604.
  • the vibrational cardiac data plotted in Figure 26 are the time and channel averaged vibrational frequenc power spectrum estimates for time slot 1 (for the symptom free person at 2630) data previously shown in Figure 21 and the person whose coronary arteries indicate coronary artery turbulence at 2620, which are data previously shown in Figure 23 and Figure 24 (2210) for the clinically diagnosed individual.
  • the roll-off of the low frequency plateau differs between the symptom free person's measurement 2620 and the clinically diagnosed person's measurement 2630.
  • the roll-off is 24 dB indicated at 2612.
  • the roil-off is 1 ? dB indicated at 2614.
  • the spread spectrum energy swath bandwidths are 90 Hz for 2606, 1 10 Hz for 2608, and 110 Hz for 2610.
  • Spread spectrum energy swath 2606 has a center frequency of 230 Hz at 2 16; spread spectrum energy swath 2 1 has a center frequency of 570 Hz at 2 18.
  • the signal-to-noise ratio for spread, spectrum swath 2606 is 1 dB and the. spread spectrum swath 2610 has a signal-to-noise ratio of 9 dB.
  • Figure 27 illustrates, generally at 2700, a method for identifying a feature related to coronary artery blood, flow turbulence using multiple humans, according to embodiments of the invention.
  • a process starts at a block 2702.
  • a time to frequency transformation is performed on vibrational cardiac data collected during a diastohc interval of a heart cycle, thereby resulting in a vibrational frequency power spectrum estimate.
  • a feature(s) is extracted from the vibrational frequency power spectrum estimate with the aid of previously identified and clinically verified features that are related to blood flow turbulence in a coronary artery and the related condition of health of the coronary artery.
  • process block 2706 includes a plurality of measurements, which can be referred to as a library of measurements.
  • the library of measurements thus created contains different features (with their associated characteristics) of blood flow turbulence in the frequency domain that are associated with different conditions of health of coronary arteries in clinically verified cases. These features are also referred to as “signatures " of blood flow turbulence and the associated pathology. Note that blood flow turbulence is not always associated with a diseased condition. At times, a measure of blood flow turbulence indicates a healthy condition (refer to Figure 19 and Figure 21), Note that in different embodiments, databases of library
  • a database can be based on transients that occur during diastole.
  • a database(s) can be assembled based on symptom free individuals, thereby establishing a plurality of measurements of healthy conditions.
  • a database(s) can be assembled that is based on features and their characteristics that are associated with clinically verified known pathologies, i.e., degree of blockage of an artery's cross -section, presence of a stent or Sack thereof, etc.
  • Databases can be combined to form databases that combine self-evaluation changes in spectral content from one time slot to another time slot within a diastolic interval along with comparison measurements from data collected on a population of humans.
  • An apparatus for performing the operations herein can implement the present invention.
  • This apparatus may be specially constructed for the required purposes, or it may comprise a general-purpose computer, selectively activated or reconfigured by a computet" program stored in the computer.
  • a computer program may be stored in a computer readable storage medium, such as, but not limited to, any type of disk including floppy disks, hard disks, optical disks, compact disk read-only memories (CD-ROMs), and magnetic-optical disks, read-only memories (ROMs), random access memories CRAMs), electrically programmable read-only memories (EPROM)s, electrically erasable programmable read-only memories (EEPROMs), FLASH memories, magnetic or optical cards, etc., or any type of media suitable for storing electronic instructions either local to the computer or remote to the computer,
  • DSP digital signal processing
  • the methods of the invention may be implemented using computer software. If written in a programming language conforming to a recognized standard, sequences of instructions designed to implement the methods can be compiled for execution on a variety of hardware platforms and for interface to a variety of operating systems, in addition, the present invention is not described with reference to any particular programming language. It will be appreciated that a variety of program ming languages may be used to implement the teachings of the inventi on as described herein. Furthermore, i t is common in the art to speak of software, in one form or another (e.g., program, procedure, application, driver,...), as taking an action or causing a result.
  • a machine-readable medium is understood to include any mechanism for storing or transmitting information in a form readable by a machine (e.g., a computer).
  • a machine-readable medium includes read only memory (ROM); random, access memory (RAM ' ); magnetic disk storage media; optical storage media; flash memo ' devices; electrical, optical, acoustical or other form of
  • propagated signals e.g., carrier waves, infrared signals, digital signals, etc.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Pathology (AREA)
  • Surgery (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Veterinary Medicine (AREA)
  • Animal Behavior & Ethology (AREA)
  • Physics & Mathematics (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Physiology (AREA)
  • Cardiology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Signal Processing (AREA)
  • Psychiatry (AREA)
  • Databases & Information Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Epidemiology (AREA)
  • Artificial Intelligence (AREA)
  • Primary Health Care (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Vascular Medicine (AREA)
  • Hematology (AREA)
  • Measuring Pulse, Heart Rate, Blood Pressure Or Blood Flow (AREA)

Abstract

L'invention concerne des procédés et des appareils pour obtenir des données cardiaques, qui incluent l'acquisition de données de champ vibratoire cardiaque à partir d'un transducteur où le transducteur mesure la vibration sur une surface d'un corps humain. Un événement coronaire indésirable est séparé des données cardiaques vibratoires. Un événement transitoire est extrait des données cardiaques vibratoires du cycle cardiaque. L'événement transitoire apparaît durant un intervalle diastolique dans un cycle cardiaque. L'événement transitoire est évalué pour un état de turbulence de flux sanguin d'artère coronaire et un état de santé d'une artère coronaire est évalué à partir d'une caractéristique dans une estimation du spectre de puissance fréquentiel vibratoire.
PCT/US2014/029833 2013-03-18 2014-03-14 Analyse spectrale de flux sanguin turbulent d'artère coronaire WO2014153265A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP14770732.7A EP2976016A4 (fr) 2013-03-18 2014-03-14 Analyse spectrale de flux sanguin turbulent d'artère coronaire
CA2907400A CA2907400A1 (fr) 2013-03-18 2014-03-14 Analyse spectrale de flux sanguin turbulent d'artere coronaire

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US13/815,961 2013-03-18
US13/815,961 US8961427B2 (en) 2008-08-09 2013-03-18 Spectrum analysis of coronary artery turbulent blood flow

Publications (1)

Publication Number Publication Date
WO2014153265A1 true WO2014153265A1 (fr) 2014-09-25

Family

ID=51581468

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2014/029833 WO2014153265A1 (fr) 2013-03-18 2014-03-14 Analyse spectrale de flux sanguin turbulent d'artère coronaire

Country Status (3)

Country Link
EP (1) EP2976016A4 (fr)
CA (1) CA2907400A1 (fr)
WO (1) WO2014153265A1 (fr)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11240579B2 (en) 2020-05-08 2022-02-01 Level 42 Ai Sensor systems and methods for characterizing health conditions

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1991006243A1 (fr) * 1989-10-26 1991-05-16 Rutgers, The State University Of New Jersey System de diagnostic non invasif pour les maladies de l'artere coronaire
US20040143191A1 (en) * 2002-11-29 2004-07-22 Yves Faisandier Device for noninvasive measurement of the blood pressure, in particular for the continuous monitoring of ambulatory blood pressure for an ambulatory patient
JP2010187928A (ja) * 2009-02-18 2010-09-02 Gifu Univ 測定対象血管の力学的機能の評価方法、測定対象血管の力学的機能評価装置、測定対象血管の力学的機能の評価プログラム及び記憶媒体
US20100317979A1 (en) * 2002-10-17 2010-12-16 Kelly Craig A Methods for Non-Invasively Monitoring Health
US20120004564A1 (en) * 2005-02-07 2012-01-05 Cardiosync, Inc. Devices and method for accelerometer-based characterization of cardiac synchrony and dyssynchrony

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5337752A (en) * 1992-05-21 1994-08-16 Mcg International, Inc. System for simultaneously producing and synchronizing spectral patterns of heart sounds and an ECG signal
EP1808122A3 (fr) * 1997-11-10 2007-08-08 Harris Corporation Système d'imagerie de flux sanguin turbulent non invasif
CN100548224C (zh) * 2004-08-31 2009-10-14 华盛顿州大学 利用超声检测内部狭窄以识别由其引起的组织振动的设备
US20070055151A1 (en) * 2005-01-20 2007-03-08 Shertukde Hemchandra M Apparatus and methods for acoustic diagnosis
US7558622B2 (en) * 2006-05-24 2009-07-07 Bao Tran Mesh network stroke monitoring appliance
US8177721B2 (en) * 2008-06-02 2012-05-15 The United States Of America As Represented By The Secretary Of The Navy Remote blood pressure waveform sensing method and apparatus
US8419651B2 (en) * 2008-08-09 2013-04-16 PhonoFlow Medical, LLC Spectrum analysis of coronary artery turbulent blood flow
EP2462871A1 (fr) * 2010-12-13 2012-06-13 Acarix A/S Système, stéthoscope et procédé pour indiquer le risque de maladie coronarienne

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1991006243A1 (fr) * 1989-10-26 1991-05-16 Rutgers, The State University Of New Jersey System de diagnostic non invasif pour les maladies de l'artere coronaire
US20100317979A1 (en) * 2002-10-17 2010-12-16 Kelly Craig A Methods for Non-Invasively Monitoring Health
US20040143191A1 (en) * 2002-11-29 2004-07-22 Yves Faisandier Device for noninvasive measurement of the blood pressure, in particular for the continuous monitoring of ambulatory blood pressure for an ambulatory patient
US20120004564A1 (en) * 2005-02-07 2012-01-05 Cardiosync, Inc. Devices and method for accelerometer-based characterization of cardiac synchrony and dyssynchrony
JP2010187928A (ja) * 2009-02-18 2010-09-02 Gifu Univ 測定対象血管の力学的機能の評価方法、測定対象血管の力学的機能評価装置、測定対象血管の力学的機能の評価プログラム及び記憶媒体

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP2976016A4 *

Also Published As

Publication number Publication date
CA2907400A1 (fr) 2014-09-25
EP2976016A1 (fr) 2016-01-27
EP2976016A4 (fr) 2016-12-07

Similar Documents

Publication Publication Date Title
US9591972B1 (en) System and method for evaluating a state of health of a coronary artery
US8419651B2 (en) Spectrum analysis of coronary artery turbulent blood flow
EP1105812B1 (fr) Systeme et procede pour faciliter la prise de decision clinique
US8684943B2 (en) Multi parametric classfication of cardiovascular sound
US8235912B2 (en) Segmenting a cardiac acoustic signal
CN103313662B (zh) 指示冠状动脉疾病的风险的系统、听诊器
CN104144636B (zh) 处理表示生理节律的信号的方法
Nigam et al. Accessing heart dynamics to estimate durations of heart sounds
CA2464029A1 (fr) Moniteur d'aeration non invasif
Kamson et al. Multi-centroid diastolic duration distribution based HSMM for heart sound segmentation
US12161449B1 (en) Systems, apparatuses, and methods for locating blood flow turbulence in the cardiovascular system
Johnson et al. Detecting aortic valve-induced abnormal flow with seismocardiography and cardiac MRI
Balogh et al. Application of phonocardiography on preterm infants with patent ductus arteriosus
Fontecave-Jallon et al. Is there an optimal localization of cardio-microphone sensors for phonocardiogram analysis?
Sæderup et al. Estimation of the second heart sound split using windowed sinusoidal models
WO2014153265A1 (fr) Analyse spectrale de flux sanguin turbulent d'artère coronaire
JP2003250767A (ja) 循環動態測定装置
Coats et al. Validation of the beat to beat measurement of blood velocity in the human ascending aorta by a new high temporal resolution Doppler ultrasound spectral analyser
CN119947653A (zh) 用于根据心音非侵入性检测肺动脉高压的可解释深度学习方法
KR101295072B1 (ko) 심플리시티 기반의 심음 분석 장치 및 그 방법
WO2023233667A1 (fr) Dispositif de traitement d'informations, procédé de traitement d'informations, système de traitement d'informations et programme de traitement d'informations
Guedes et al. Heart sound analysis for blood pressure estimation
Thibbotuwawa Gamage Seismocardiography-Genesis, and Utilization of Machine Learning for Variability Reduction and Improved Cardiac Health Monitoring
US12023165B2 (en) Screening cardiac conditions using cardiac vibrational energy spectral heat maps
Kribèche et al. Detection and analysis of fetal movements by ultrasonic multi-sensor Doppler (ACTIFOETUS)

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: 14770732

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2014770732

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 2907400

Country of ref document: CA

NENP Non-entry into the national phase

Ref country code: DE

点击 这是indexloc提供的php浏览器服务，不要输入任何密码和下载