U.S. patent application number 10/303252 was filed with the patent office on 2003-06-12 for method and apparatus for measuring volume flow and area for a dynamic orifice.
This patent application is currently assigned to The General Hospital Corporation. Invention is credited to Buck, Thomas, Levine, Robert A., Mucci, Ronald A..
Application Number | 20030109785 10/303252 |
Document ID | / |
Family ID | 26821019 |
Filed Date | 2003-06-12 |
United States Patent
Application |
20030109785 |
Kind Code |
A1 |
Buck, Thomas ; et
al. |
June 12, 2003 |
Method and apparatus for measuring volume flow and area for a
dynamic orifice
Abstract
Techniques are provided for obtaining selected instantaneous
area measurements for a dynamic orifice through which blood is
flowing in at least one direction, for obtaining instantaneous flow
rates of blood passing through such a dynamic orifice and for
obtaining flow volume for blood passing through a dynamic orifice.
All of these techniques involve ensonifying a thin sample volume of
blood flow exiting at the orifice, which volume is in a region of
flow which is substantially laminar, such region normally being the
vena contracta for the orifice, with an ultrasonic pulsed Doppler
signal, receiving backscattered signal from blood within the sample
volume and forming a power-velocity spectrum from the received
backscattered signal. Techniques are disclosed for assuring that
only laminar flow is looked at in forming the power-velocity
spectrum. To obtain instantaneous area measurements, the power
integral of laminar flow from the spectrum is formed, this power
integral being proportional to an instantaneous cross-sectional
area of the orifice. For instantaneous flow rates, the
instantaneous power-velocity integral is formed from the laminar
flow of the spectrum, while flow volume is obtained from the timed
integral of the instantaneous flow rate.
Inventors: |
Buck, Thomas; (Cambridge,
MA) ; Levine, Robert A.; (Brookline, MA) ;
Mucci, Ronald A.; (Westwood, MA) |
Correspondence
Address: |
Peter C. Lando
Wolf, Greenfield & Sacks, P.C.
600 Atlantic Avenue
Boston
MA
02210
US
|
Assignee: |
The General Hospital
Corporation
Boston
MA
|
Family ID: |
26821019 |
Appl. No.: |
10/303252 |
Filed: |
November 25, 2002 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
10303252 |
Nov 25, 2002 |
|
|
|
09467165 |
Dec 20, 1999 |
|
|
|
6544181 |
|
|
|
|
60122926 |
Mar 5, 1999 |
|
|
|
Current U.S.
Class: |
600/437 ;
600/465 |
Current CPC
Class: |
A61B 8/488 20130101;
G01S 7/5205 20130101; A61B 8/0883 20130101; G01F 1/663 20130101;
G01S 15/8979 20130101; A61B 8/06 20130101 |
Class at
Publication: |
600/437 ;
600/465 |
International
Class: |
A61B 008/00; A61B
008/14 |
Goverment Interests
[0002] Work in this invention was supported in part by Grants
HL38176 and HL57302 from the National Institute of Health,
Bethesda, Md. and BU 1097/-1 from the Deutsche
Forschungsgemeinschaft, Bonn, Germany.
Claims
1. A method comprising: ensonifying a sample volume of a fluid
exiting an orifice with an ultrasonic signal; receiving a
backscattered signal from the sample volume; and forming a
power-velocity spectrum from the backscattered signal; and forming
an integral for the power-velocity spectrum that is proportional to
a cross-sectional area of the orifice.
2. The method of claim 1, wherein the integral is a power
integral.
3. The method of claim 1, wherein the integral is a time integral.
Description
RELATED APPLICATION
[0001] This application claims the benefit of co-pending patent
application Ser. No. 09/467,165, filed Dec. 20, 1999, which
application claims priority from Provisional Application Serial No.
60/122,926, filed Mar. 5, 1999.
FIELD OF THE INVENTION
[0003] This invention relates to methods and apparatus for
utilizing an ultrasonic pulsed wave Doppler signal to measure the
instantaneous area of a dynamic orifice through which blood is
passing and/or to measure instantaneous flow rate and flow volume
of blood passing through such a dynamic orifice, and more
particularly to such methods and apparatus which involves
ensonifying a sample volume of blood flow exiting the orifice and
identifying the region of such flow which is substantially
laminar.
BACKGROUND OF THE INVENTION
[0004] Ultrasound, and more specifically the frequency range from 1
MHz to 5 MHz, is used for real-time imaging of the beating heart.
In the human heart, the efficiency of getting blood pumped through
the body is dependent on a series of four one-way valves, each
separating the two contracting chambers of the heart, which valves
are prone to a variety of diseases, often times resulting in their
inability to close properly. Ultrasound, through the use of the
Doppler concept, is able to obtain information pertaining to blood
flow within the heart and in the vicinity of the valves for
diagnostic purposes, ultrasound having become the most important
noninvasive diagnostic technique for cardiovascular disease.
However, the use of noninvasive ultrasound techniques to quantify
pathologic backflow associated with valvular heart disease, other
cardiac pathologies such as inter-septal shunting and other blood
flows through dynamic orifices of unknown area has been an elusive
medical goal for many years.
[0005] While for the purpose of this discussion, focus will be on
valvular regurgitation, that is the pathologic backflow of blood
through a one-way valve when in the closed state, which is a
serious, and at times life-threatening, condition common in
virtually all acquired and congenital heart disease, the invention
is by no means limited to this application, and some other
applications will be discussed later.
[0006] Leakage of one or more valves is caused by various diseases
which prevent the leaflets of the valve from closing sufficiently,
thereby creating a lesion called a regurgitant orifice. There is a
need to accurately measure the volume of regurgitation (reverse
blood flow) as a guide both to diagnosis and to therapy, especially
now that valve repair techniques allow interventions to be
considered earlier in the disease before dilation of the chambers
(atria and ventricles) and subsequent heart failure occur. Current
uncertainties regarding the natural history of the valve disease
and the optimal timing of surgery are compounded by a limited
ability to measure the basic lesion. Noninvasive procedures for
quantification of regurgitant volume based on ultrasound do exist,
but are subject to limitations that include: inaccurate diagnosis
of lesion severity resulting from indirect measurements, multiple
step procedures prone to error, and limiting assumptions about the
flow associated with the lesion. In fact, there is currently no
truly satisfactory method for noninvasive quantification, and even
routine invasive methods, being costly and potentially risky, are
only semi-quantitative. Those invasive methods are based on direct
catheterization of the heart that allows obtaining information
about flow, volume, pressure, etc.
[0007] The fundamental problem in using noninvasive ultrasound is
that Doppler measures the velocity, not the desired volume, of
regurgitant blood flow. Therefore, in order to determine volume of
blood passing through an orifice, for example the regurgitant
orifice of a diseased heart valve, the area of flow, also referred
to as the effective orifice area, has to be known. All methods to
date have failed to measure the effective orifice area accurately
because of the complex shape and dynamic changes of this area
throughout the period of flow.
[0008] A potential solution is to use the backscattered acoustic
power measurements of the received spectral Doppler signal as a
measure of the area of flow. It is well known that each frequency
component of the Doppler spectrum provides a measurement of
acoustic power that is proportional to the volume of scatterers
moving through the Doppler ultrasound beam at the velocity
corresponding to the Doppler frequency. It follows that velocity
times power, integrated over the entire velocity spectrum, should
then be proportional to the volume flow rate Q.cndot. of all
scatterers (mainly red blood cells) passing through the ultrasound
beam, since the blood volume is related to the concentration of red
blood cells by way of the hematocrit.
[0009] This Doppler power principle holds only for laminar flow and
was applied to flow in vessels but it has long been assumed that it
cannot be applied to regurgitant jets, that is jets comprised of
the regurgitant flow of blood, since the assumption is that the jet
contains turbulent eddies which are believed to increase the
backscattered power. In addition, entrainment of blood into the jet
can contribute to the overestimation of the actual flow through the
orifice.
[0010] While the problems of measuring flow volume and/or orifice
area for a dynamic orifice through which blood flows is a
particular problem when measuring regurgitant flow through a heart
valve, similar problems arise in measuring valvular stenosis,
septal defects with shunt flow, and peripheral vascular disease
with vessel obstruction. In these and other applications, a need
exists for an improved noninvasive method and apparatus for
measuring flow volume and/or orifice area for a dynamic orifice
having blood flow therethrough, which technique does not suffer the
limitations discussed above for existing methodologies.
SUMMARY OF THE INVENTION
[0011] In accordance with the above, a method and apparatus are
provided for obtaining instantaneous area measurements for a
dynamic orifice through which blood is flowing in at least one
direction. The technique involves ensonifying a thin sample volume
of blood flow exiting the orifice, which volume is in a region of
flow which is substantially laminar, with an ultrasonic pulsed
Doppler signal; receiving backscattered signal from blood within
the sample volume; forming a power-velocity spectrum from the
received backscattered signal; and forming the power integral of
the laminar flow from the spectrum, this power integral being
proportional to an instantaneous cross-sectional area of the
orifice. A time profile of instantaneous areas of flow for the
orifice may be obtained by repetitively performing the laminar flow
power integral measurement for successive time intervals. The
portion of laminar flow in the power velocity spectrum is
preferably determined. For preferred embodiments, the sample volume
is at the vena contracta of flow exiting the orifice. The vena
contracta is the smallest cross-sectional area traversed by flow
just beyond the orifice, it being found that flow is substantially
laminar at the vena contracta, this vena contracta being the region
where entrainment of flow turbulence is at its minimum. For
preferred embodiments, the ultrasonic Doppler signal is
electronically steered and focused to the vena contracta and is
preferably wide enough so as to fully ensonify the vena contracta.
The electronic steering and focusing may be performed by moving the
ultrasound signal through blood flow exiting the orifice, and
detecting a Doppler spectral display and/or audio output, the
signal being at the vena contracta when the Doppler signal consists
primarily of laminar flow. To assure that only signal from laminar
flow is utilized in performing the power integral calculation, the
power velocity spectrum is preferably smoothed to eliminate the
effects of any aberrations therein, and the velocity for peak power
is determined for each time interval. A lower velocity of laminar
flow is then determined as being a selected velocity, for example
the maximum velocity, which is less than the velocity at peak power
where the power is at a selected percentage of the peak power, and
an upper velocity of laminar flow is determined which is a selected
velocity, for example a minimum velocity, greater than the velocity
at peak power where the power has dropped to a specified percentage
of the peak power. Depending on application, the percentage of peak
power may be from approximately 30% to approximately 60%, with
approximately 50% or -3 db being the percentage of peak power for
an illustrative embodiment. Only the power-velocity spectrum
between the lower velocity and upper velocity, which is assumed to
be derived from flow which is substantially laminar, is utilized in
doing the power integral calculation, thus assuring that various
values determined utilizing the teachings of this invention are
obtained only from readings of laminar flow. The flow may for
example be regurgitant flow through a faulty heart valve, the
orifice area being that of lesions in the heart valve permitting
the regurgitant flow.
[0012] The technique may also include calibrating to permit
absolute flow area to be obtained. Calibrating may include applying
a narrow ultrasound reference beam placed within the laminar flow
in the vena contracts, the reference beam having a known
cross-sectional area (CSA.sub.ref), and computing flow
cross-sectional area (CSA.sub.flow) from
CSA.sub.flow=CSA.sub.ref.multidot.PI.sub.meas/PI.sub.ref, where
PI.sub.meas and PI.sub.ref equal the power measure by a broad
measurement beam encompassing the vena contracta and the power
measure by the narrow reference beam of known cross-sectional area,
respectively. Where the flow being measured is regurgitant flow
through a faulty heart valve, calibration may be performed by
detecting backscattered Doppler ultrasound power from the reference
beam for forward flow when the valve is open.
[0013] The invention also involves a technique for obtaining
instantaneous flow rates of blood passing through a dynamic orifice
in at least one direction, which technique includes ensonifying a
thin volume of blood flow exiting the orifice, which volume is in a
region of flow which is substantially laminar, with an ultrasonic
pulsed wave Doppler signal which fully encompasses the
cross-sectional area of the volume where flow is substantially
laminar; receiving backscattered signal from blood within the
pulsed wave Doppler sample volume; forming a power-velocity
spectrum from received backscattered signal; and forming the
instantaneous power-velocity integral (PVI) from the laminar flow
of the spectrum. A pulsed wave Doppler signal, such as high-PRF
Doppler, capable of representing the full range of velocity, is
preferred. A time profile of instantaneous flow rates for the
orifice may be obtained by calculating the instantaneous
power-velocity integrals for successive time intervals. As for the
flow area determination, the thin sample volume is preferably at
the vena contracta of the flow exiting the orifice, with electronic
steering preferably being performed on the ultrasonic signal to
steer and focus it to the vena contracta and is wide enough so as
to fully ensonify the vena contracta. The electronic steering and
focusing may be performed by scanning the ultrasound signal through
blood flow exiting the orifice and detecting at least one of
Doppler spectral display and audio output, the signal being at the
vena contracta when the output of the detecting step identifies
laminar flow. Flow at nonlaminar velocities may be eliminated from
the PVI determination in the manner described above. The flow may
for example be regurgitant flow through a faulty heart valve, the
orifice area being that of lesions in the heart valve permitting
regurgitant flow.
[0014] Calibration may also be performed to permit absolute flow
rate to be obtained, calibration including applying a narrow
ultrasound reference beam placed within the laminar flow in the
vena contracta, the reference beam having a known CSA
(CSA.sub.ref), and computing Flow rate from Flow
rate=CSA.sub.ref.multidot.PVI.sub.meas/PI.sub.ref. Where the flow
being measured is regurgitant flow through a faulty heart valve,
calibration may include detecting backscattered Doppler ultrasound
power from the reference beam when the valve is open for forward
flow.
[0015] Finally, the invention involves a technique for obtaining
flow volume for blood passing through a dynamic orifice in at least
one direction, which technique includes ensonifying a thin volume
of blood flow exiting the orifice, which volume is in a region of
flow which is substantially laminar, with an ultrasonic pulsed wave
Doppler signal which fully encompasses the cross-sectional area of
the volume where flow is substantially laminar; receiving
backscattered signal from blood within the pulsed wave Doppler
sample volume; forming a power-velocity spectrum from received
backscattered signal; and forming the instantaneous power-velocity
integral (PVI) from the laminar flow portion of the spectrum and
the profile of the instantaneous flow rates. The flow volume is
obtained from the time integral of the instantaneous flow rate
(PVTI). As for prior embodiments, a pulsed wave Doppler signal,
such as high-PRF Doppler, is preferred, the thin sample volume is
preferably at the vena contracta of the flow exiting the orifice,
the technique preferably includes electronic steering and focusing
the ultrasonic signal to vena contracta, and non-laminar velocities
being removed from the calculations using techniques previously
discussed. The ultrasound signal is preferably wide enough so as to
fully ensonify the vena contracta. The flow may for example be
regurgitant flow through a faulty heart valve, the orifice area
being that of lesions in the heart valve permitting regurgitant
flow. Where the flow is regurgitant flow through a faulty heart
valve, detected forward and regurgitant flow may be combined to
obtain a measure of regurgitant fraction.
[0016] Calibration may also be performed to permit absolute flow
volume to be obtained, calibration including applying a narrow
ultrasound reference beam placed within the laminar flow in the
vena contracta, the reference beam having a known CSA
(CSA.sub.ref), and computing Flow volume from Flow
volume=CSA.sub.ref.multidot.PVTI.sub.meas/PI.sub.ref. Where the
flow being measured is regurgitant flow through a faulty heart
valve, calibration may include detecting backscattered Doppler
ultrasound power from the reference beam when the valve is open for
forward flow.
[0017] The foregoing and other objects, features and advantages of
the invention will be apparent from the following more particular
description of preferred embodiments of the invention as
illustrated in the accompanying drawings.
IN THE DRAWINGS
[0018] FIG. 1A is a representation of the anatomy and physiology of
the human heart, FIG. 1B and FIG. 1C being enlarged views for a
portion of the heart shown in FIG. 1A during diastolic and systolic
phases, respectively;
[0019] FIG. 2A is an illustration of data acquisition using an
ultrasonic transducer in accordance with the teachings of this
invention;
[0020] FIG. 2B is an ultrasonic display for two-dimensional image
mode, and FIG. 2C is an exemplary ultrasonic display at the vena
contracta for Doppler mode;
[0021] FIG. 3 is a representation of the anatomy for a regurgitant
jet;
[0022] FIG. 4A is a diagrammatic representation of the sonification
of an area of flow by a Doppler beam for a Doppler angle
.theta.=0;
[0023] FIG. 4B is a diagrammatic representation of an exemplary
velocity distribution across an area of flow;
[0024] FIG. 4C is a plot of backscattered power versus frequency
from the received Doppler spectrum shown in FIG. 4B
[0025] FIG. 4D is a plot illustrating the conversion of power and
frequency to display video intensity and velocity;
[0026] FIGS. 4E and 4F are exemplary displays of actual flow
measurements (velocity vs. time) by Doppler for in vitro steady
flow and in vivo parabolic flow, respectively;
[0027] FIG. 5 is a diagrammatic illustration of the principle of
volume flow rate calculation over an area of flow for a spectrum of
three different flow velocities;
[0028] FIG. 6 is a diagrammatic representation illustrating the
proportionality of backscattered Doppler power to a sonified volume
of blood;
[0029] FIG. 7 is a diagram illustrating the independence of the
power-velocity spectrum from the distribution of velocities across
the flow CSA;
[0030] FIG. 8 is a diagrammatic representation illustrating
calibration using a broad measurement beam and a narrow reference
beam;
[0031] FIGS. 9A-9D are a high-PRF Doppler signal, an enlarged
sector of the signal shown in FIG. 9A, a PI plot and a PVI plot,
respectively for an exemplary orifice and flow profile;
[0032] FIG. 10 is a schematic representation of the flow phantom
used for in vitro experiments for validation the teaching of this
invention;
[0033] FIGS. 11A and 11B are plots of various results for PI
achieved in an in vitro study of this invention;
[0034] FIG. 12 is a plot of various calculated and calibrated PVI
values versus actual flow rates for in vitro experiments involving
the teaching of this invention;
[0035] FIG. 13 is a plot of calculated and calibrated PVTI values
versus actual regurgitant flow volumes for exemplary in vitro
experiments involving the teaching of this invention;
[0036] FIG. 14 is a schematic diagram of an in vivo animal model
used in performing in vivo experiments involving this
invention;
[0037] FIG. 15 is a plot of calculated and calibrated mitral
regurgitant stroke volume (MRSV) versus actual values of MRSV for
in vivo animal experiments;
[0038] FIGS. 16A-16C are illustrative Doppler spectra for a patient
with functional mitral regurgitation;
[0039] FIG. 17 is a plot of calculated and calibrated mitral
regurgitant stroke volume (MRSV) versus independent reference
values of MRSV in patient studies;
[0040] FIG. 18A is a schematic structure presentation, and FIG. 18B
are plots useful in conjunction with Appendix A for describing the
principles of pulse Doppler for blood velocity measurement;
[0041] FIG. 19 is a plot useful in conjunction with Appendix B for
illustrating beam properties, and in particular the use of a
smaller aperture to produce a larger beam CSA; and
[0042] FIG. 20 is a plot useful in conjunction with Appendix C for
understanding broadening of the power-velocity spectrum of laminar
flow.
DETAILED DESCRIPTION
[0043] Referring first to FIG. 1A, the heart is shown as having
four chambers--left and right ventricle (LV and RV) and left and
right atrium (LA and RA). These chambers are connected by four
one-way valves, the mitral (MV), aortic (AV), tricuspid (TV) and
pulmonary (PV) valves. There are also various vessels going in and
out of the heart. The arrows in FIG. 1A indicate the direction of
the blood where the venous return enters the right heart via the
right atrium and is pumped by the right ventricle via pulmonary
valve and pulmonary artery to the left and right lung. After
oxygenation in the lungs, the blood re-enters the heart via the
pulmonary veins and flows to the left ventricle via the left atrium
and the mitral valve. The left ventricle, as the strongest
contracting chamber, pumps the oxygenated blood to the entire body
via the aortic valve and the aorta, thereby creating the systemic
blood pressure.
[0044] FIGS. 1B and 1C indicate the direction of blood flow during
the two phases of one heart cycle, diastole where the left
ventricle relaxes to refill with blood from the left atrium, and
systole where the left ventricle contracts to eject the blood to
the aorta. FIG. 1C shows the pathologic regurgitant flow when the
mitral valve closes incompletely during systole.
[0045] FIG. 2A illustrates the use of an ultrasonic transducer to
acquire data on the heart. The ultrasonic transducer is positioned
onto the outer chest wall of the patient, and its beam is oriented
to the region of flow. A high-PRF Doppler sample volume is
depicted, which is located in the vena contracta which, as
indicated previously, is the narrowest portion of the jet. FIG. 2B
shows how a two-dimensional image plane intersecting the heart is
used to navigate the Doppler beam so as to place the sample volume
at the vena contracta, which is at the white arrow in FIG. 2B. Once
the beam is located, the system may be changed to pulsed Doppler
mode to register the Doppler signal as depicted in FIG. 2C. The
velocities are displayed below a baseline of 0, indicating that the
flow is away from the transducer passing from the left ventricle
through the diseased mitral valve back to the left atrium.
[0046] FIG. 3 is a diagram illustrating the anatomy of a
regurgitant jet, illustrating the vena contracts, the core of
laminar flow, and the regions of entrainment of fluid and
turbulence. The vena contracta is shown as the narrowest portion of
the regurgitant jet just below the orifice where velocities are
highest and flow is laminar across the jet prior to entrainment of
fluid and turbulence. The figure also indicates the Doppler beam
and the sample volume located at the vena contracta.
Cross-sectional views at the right indicate the laminar flow at the
vena contracta and the turbulent flow beyond the vena
contracta.
[0047] Basic Principles
[0048] Referring to FIG. 4A, for Doppler operation, the ultrasound
system, through the use of a hand-held transducer, transmits an
ultrasound beam that is electronically steered and focused. It is
the backscattered ultrasound that provides the diagnostic
information, the ultrasound energy within the beam backscattered as
it propagates within the body. To acquire the backscattered signal,
the ultrasound system operates in `reverse` using the same
transducer to receive the backscattered energy, again with
electronic steering and focusing.
[0049] The spectral Doppler power is proportional to the volume of
backscatterers on a frequency basis. This power spectrum is mapped
to a vertical line of display intensity. The spectrum is computed
and displayed continually and in real-time as depicted in FIGS. 4E
and 4F. Parabolic flow (FIG. 4F) pertains to the shape of the
narrow-band velocity trace over time from an in vivo pulsatile flow
through a mitral valve regurgitant orifice during systolic left
ventricular contraction in a patient. For FIG. 4C, power at f2 is
twice as large as at f1 or f3 because the area of V2 is twice as
large as of V1 and V3.
[0050] In the Doppler modality of operation, the power spectrum of
the demodulated received waveform is computed using an FFT
algorithm and analyzed by the ultrasound imaging system. The
analysis of the received Doppler signal provides two important
components; the Doppler frequency and the corresponding
backscattered signal strength (acoustic power).
[0051] The velocity of the backscattering medium, presumably blood
within a sample volume at the focal region, is related to the
frequency of the power spectrum of the backscattered ultrasound by
the equation, 1 f D = 2 f c | V SC | cos ( 1 )
[0052] where f.sub.D denotes the Doppler frequency of the
backscattered signal, f denotes the transmit frequency of the
ultrasound, c denotes the speed of propagation of sound within the
medium, (the approximate speed of sound in water is 1470 m/sec, in
soft tissue 1540 m/sec and in bone 4800 m/sec), V.sub.SC denotes
the velocity of the scattering object, presumably blood, and
.theta. denotes the angle between the direction of the ultrasound
beam and the direction of blood velocity. A derivation of Eq. 1
based on the amplitude/phase data of the demodulated received
signal rather than the basic Doppler concept is provided in
Appendix A.
[0053] Using "somewhat conventional" procedures, for example the
use of the fast Fourier transform (FFT), the spectral content of
the backscattered Doppler signal is determined. The spectral bins
of sufficient amplitude not only indicate blood flow at velocities
as determined by Eq. 1, but also backscattered power is
proportional to the amount of blood flowing at the corresponding
velocity (FIG. 4C). However, in the present systems, where the
Doppler spectrum of blood flow is generated and displayed in
real-time, only the velocities are interpreted for diagnostic
purposes.
[0054] The basic concept for using the Doppler spectral information
to quantitate flow rate and volume is as follows:
[0055] First, referring to FIG. 5, for the purpose of concept
presentation a lumen is defined, as a bounded cross-sectional
planar area denoted by the vector A.sub.l.fwdarw. of direction
perpendicular to the planar surface, through which flows an
incompressible fluid with directional velocity function
v.fwdarw.(x, y, t) across the area of flow at instantaneous time t.
The incremental flow .delta.Q.cndot.(x, y, t) passing through the
differential area .delta.A.fwdarw.(x, y) at time t is given by
v.fwdarw.(x, y, t).cndot..delta.A.fwdarw.(x, y) where .cndot.
denotes the dot product operation of two direction vectors. The dot
product of two vectors a and b is defined as the product of their
magnitudes times the cosine of the angle between them. The total
flow passing through the planar cross-sectional area encompassed by
the lumen is given by the integral expression 2 Q ( t ) = A i v ( x
, y , t ) A ( x , y ) ( 2 )
[0056] The flow volume, Q.sub.v, within an interval of time, for
example, within the systolic portion of the cardiac cycle, can be
obtained by integrating the instantaneous flow over the time
interval of interest denoted T, that is, 3 Q v = T Q v ( t ) t ( 3
)
[0057] The uniqueness of this approach is the capability to measure
the regurgitant flow rate and flow volume noninvasively and
directly at the valve lesion using only the backscattered power and
velocity information, both of which are contained in the received
ultrasound spectrum. This is accomplished as follows: A sample
volume of cross-sectional area that encompasses the area of
regurgitant flow at the vena contracta is sonified with the
ultrasound beam. It is important to note that the flow within the
sample volume when placed at the vena contracta contains the
laminar flow of the vena contracta and nonlaminar flow of the blood
surrounding the vena contracta unlike in prior art where blood flow
was constrained within the wall of the vessel. While the Doppler
spectrum from a sample volume applied to such a vessel only
contains the signal from the vessel flow, the Doppler spectrum from
the sample volume of a regurgitant jet at the level of the vena
contracta contains signal from laminar and turbulent flow. This
requires the identification of the laminar portion of the Doppler
spectrum and the use only of this portion in the calculation of the
measurements.
[0058] To accomplish the above, the operator of the system
initially utilizes a standard ultrasonic beam control, such as a
joy stick or tracker ball, to adjust the beam as shown in FIG. 2A
and 2B to pass through the mitral valve or other valve of interest.
The depth for imaging is then adjusted on the machine in standard
fashion so as to be directly adjacent a regurgitant side of the
valve at the vena contracta. All this is done with the ultrasound
system in image mode as shown in FIG. 2B, the desired imaging spot
being generally at the bright dot being pointed to by the
arrow.
[0059] At this point, the beam is focused at substantially the
desired point, but may not be exactly centered with the flow to be
measured. To fine tune the position, the ultrasound system is
switched to Doppler mode, resulting in a display such as that shown
in FIG. 2C, illustrating velocity versus time. FIG. 9A is an
alternative image. The bright signals in the low velocity area on
these displays, for example under 100, is generally tissue noise,
turbulent flow, and the like, while the bright higher frequency
band on the outside of each of these images is generally from
laminar flow. When in the vena contracta where laminar flow
predominates, the narrowness and cleanness of this bright
narrow-band spectrum will be optimized. Generally, in order to
eliminate the effect of aberrations in the backscattered signal,
the backscattered signal is smoothed utilizing a standard averaging
or smoothing technique. The operator then looks at the Doppler mode
image and fine tunes transducer position and/or image depth until
the "cleanest" observed image is obtained. Alternatively, the
smoothed backscattered signal may be converted to audio, the signal
from the vena contracta being the cleanest tone involving the
fewest frequencies. Theoretically this fine tuning could be done
automatically, with for example the beam position and depth being
slightly adjusted through a small range, and for example an audio
signal being monitored to provide a feedback signal for controlling
the beam in a direction resulting in the cleanest audio output.
This could be done first for position and then for depth, or the
two could be done simultaneously in accordance with a selected
feedback controlled scanning algorithm.
[0060] However, as indicated above, even when the beam is
positioned exactly at the vena contracta, in order for the beam to
ensonify the entire vena contracta, it will inherently also
ensonify some of the surrounding turbulent flow. In order to ensure
that only signal from laminar flow at the vena contracta is
utilized for performing the various calculations in accordance with
the teachings of this invention, a velocity filtering technique may
be utilized, which technique is illustrated in FIGS. 9A and 9B.
Referring to FIG. 9B, which is an enlarged image of the portion of
FIG. 9A shown in the rectangle, it is seen that the backscattered
signal contains a number of velocities, the power at a given
velocity for the time t.sub.i being indicated by the brightness of
the image for the velocity V.sub.1 at the given time, a brighter
image indicating more power. For each time t.sub.i a determination
is made of the velocity of peak power in the smoothed
power-velocity spectrum for the sample t.sub.l in FIG. 9B, which is
also shown at the right of the Figure. From this image, peak power
is found to extend over a velocity range V.sub.r, a velocity in
this range being utilized as the velocity of peak power for this
time interval. A determination is then made of the lower velocity
of laminar flow (V.sub.low) which is considered to be the maximum
velocity which is less than the velocity for peak power (i.e.,
V.sub.R) where the power is a specified percentage of peak power.
This percentage may, for example, be in an approximate range of 30%
to 60% of peak power depending on application, being 50% or -3 DB
for an illustrative embodiment. Similarly V.sub.high, the upper
velocity of laminar flow is determined by finding the minimum
velocity which is greater than the velocity V.sub.R of peak power
where the power is the specified percentage of the peak power.
While the specified percentage of peak power need not be the same
for determining V.sub.low and V.sub.high, these percentages are the
same for an illustrative embodiment and would normally be the same.
Once V.sub.low and V.sub.high are determined for a given time
t.sub.I, only backscattered radiation having a velocity falling
within this range is utilized and performing the various
calculations, thus assuring that only laminar flow at the vena
contracta is being utilized for these calculations.
[0061] Pulse-wave (PW) Doppler is used since, as described in
Appendix A, this allows a range gate to be set which in turn
controls the thickness of the interrogated volume. In PW Doppler, a
relatively short pulse is transmitted and the sample volume
thickness is controlled by receiving and analyzing the
backscattered ultrasound for a relatively short interval of time.
The total backscattered acoustic power from such a Doppler sample
volume is linearly proportional to the sonified blood volume, where
the proportionality factor is given by the backscattering
coefficient that is related to the number of independent
scatterers, mainly red blood cells (erythrocytes), in this
volume.
[0062] Further, if a disk-like sample volume is traversed by blood
flow, then the backscattered power is linearly proportional to the
cross-sectional area (CSA) of the flow within the lumen projected
onto the surface of the sample volume of the sonifying ultrasound
beam as shown in FIG. 6. This assumes that the thickness of the
sample volume is thin and constant, that the concentration of
scatterers (hematocrit) is constant, and that the beam encompasses
the entire CSA of flow. For example, the total backscattered power
from a jet with a CSA of 1.0 cm.sup.2 would be twice as much as
from a jet with a jet CSA of 0.5 cm.sup.2.
[0063] This approach does not require the existence of a lumen nor
the explicit determination of the velocity distribution over the
cross-sectional area of the vena contracta. Rather, the Doppler
beam encompassing the entire area of flow, effectively integrates
the contribution of all scatterers at each velocity of laminar flow
regardless of the velocity profile (FIG. 7). That is, in the
Doppler modality the backscattered acoustic signal power increases
linearly by the number of scatterers traveling at the same
velocity, independent of their individual locations within the jet
CSA and independent of their individual locations within the sample
volume. As a consequence, it becomes clear that with this
application of the Doppler beam analyzing the backscattered
power-velocity spectra, there are no longer concerns about the
velocity distribution of scatterers across the flow CSA or about
the shape and area of the flow CSA (FIG. 7), assuming a uniform
sensitivity across the transmit and receive beam CSA. Thus, in FIG.
7, the areas of V1, V2, and V3 in flows A and B having the same
size but different shapes are represented by the same power
spectrum.
[0064] In order to detect the weak backscattered signal from blood,
the Doppler modality requires the elimination of signal
backscattered from tissue prior to the calculation of the power
spectrum, since the backscattered power of tissue can be 100 to
1000 times greater than the power backscattered by blood. This is
done by a high-pass filtering process, referred to as a wall
filter, that eliminates Doppler signals backscattered from low
velocity targets such as tissue. Hence even though the
backscattered power is independent of velocity, only the
backscattered power of blood flow above the cutoff velocity of the
wall filter is measured. From this it also becomes clear that blood
within the sample volume that is not flowing, such as that external
to the vena contracta, does not contribute to the power spectrum
and therefore does not contribute to the area and associated flow
measurement.
[0065] In summary, for a sample volume of uniform thickness,
achieved with PW Doppler and placed in the vena contracta, the
spectral Doppler power associated with a Doppler frequency, denoted
P(v), is linearly proportional to the cross-sectional area of
scatterers traveling at the velocity v, corresponding to the
Doppler frequency. Hence, the total spectral power is linearly
proportional to the cross-sectional area of the vena contracta.
That is 4 Power = vel P ( v ) v CSA jet ( 4 )
[0066] This is demonstrated in the results that are contained
herein.
[0067] Furthermore, the power associated with a given frequency (or
velocity) times the velocity is equivalent to the area times the
velocity, which in turn provides a measure of the component of flow
attributable to scatterers of the specified velocity. If this
calculation of power times velocity is integrated for each
component of the Doppler spectrum, an estimate proportional to the
total instantaneous flow passing through the vena contracta,
denoted Q.cndot..sub.v(t), can be obtained. That is, 5 Q ( t ) vel
P ( v ) v v ( 5 )
[0068] The flow volume can be obtained from the estimates of
instantaneous flow by integrating over the time interval of
interest denoted T. That is, 6 Q v = T Q ( t ) ( 6 )
[0069] These relationships are demonstrated in both in vitro and in
vivo experiments to be discussed later.
[0070] From this discussion, it is seen that the Doppler
power-velocity principle also indicates that the backscattered
power integrated over the velocity spectrum at the vena contracta
is proportional to the cross-sectional area of the vena contracta,
regardless of flow rate. Of note is that the proportionality
between power and CSA depends mainly on attenuation and the
backscattering coefficient, which is a nonlinear function of
hematocrit. The calibration technique described later, however,
takes this into account by comparing measured power over the entire
sample volume with that in a small reference beam of known CSA.
[0071] Finally, although there is a dependence of the velocity
measurement on the angle .theta. between the direction of flow in
the vena contracta and the main axis of the Doppler beam as
indicated in Eq. 1, the measurement of power times velocity is
independent of this angle. That is, the Doppler velocity is
decreased by the cos .theta. and the power measurement, being
proportional to the CSA of flow projected onto the surface of the
sample volume, is increased by 1/cos .theta., such that 7 Flow rate
= ( v cos ) ( Flow CSA cos ) ( 7 )
[0072] This angle independence becomes relevant in in vivo studies
where the Doppler beam cannot always be aligned with the direction
of flow.
[0073] In the following section, evidence is discussed establishing
that the regurgitant flow volume can be estimated/quantified using
only the power and velocity information acquired with Doppler
ultrasound, the discussion including both methodology of the
conducted experiments and the results.
[0074] Methodology
[0075] Procedures required in practicing the teaching of this
invention included: 1) generation of a Doppler beam sufficiently
wide to encompass the cross-sectional area of the vena contracta of
a regurgitant jet, and 2) a method of calibration to compensate for
variations in attenuation and the backscattering coefficient. The
following is a brief description of each procedure:
[0076] Doppler Beam-Width
[0077] For an illustrative embodiment, the full 1.2 cm by 2.0 cm
transducer aperture of the transducer used produces a beam having a
cross-sectional area of approximately 3.1 mm in elevation by 5.2 mm
in the lateral dimension at a 10 cm depth, based on defining the
border of the beam CSA by where the sensitivity of the transducer
drops to the half-maximal power or -3 db. To ensonify larger flow
areas, a broader measurement was produced based on the principle
that the smaller the transducer aperture, the broader the distal
beam (see FIG. 8). The aperture for the left measurement beam was
diminished with the application of a Tyvek (Dupont) mask over the
transducer face. In vitro, a 7-mm diameter circular aperture was
applied, thereby increasing the half-maximum-power beam-width at a
10-cm depth to approximately 6.75 mm (CSA=0.36 cm.sup.2; Appendix
B). However, it was necessary to use a 10-mm aperture creating a
beam-width of 5.8 cm (CSA=0.26 cm.sup.2) at 10 cm for the in vivo
experiments to minimize the loss of transmit power and Doppler
receive sensitivity resulting from decreasing the size of the
transducer aperture without compensating for the decrease of
transmit power. Alternative beam broadening techniques are
discussed in Appendix D.
[0078] Calibration of Doppler Power Measurements
[0079] Although the invention only requires that a proportionality
exist between PVI and flow, it is necessary to calibrate the system
in order to obtain absolute measurements of area and flow rate in a
clinically useful manner. That is, in part because the power
measurements obtained by the ultrasound system are unitless, and in
part because for the same blood volume, Doppler power measurements
will vary among patients due to differences in attenuation and
backscattering coefficients (related to patient hematocrit). The
Doppler measurements in each individual can be calibrated with a
narrow reference beam that fits within an area of flow in the
vicinity of the vena contracta (right beam FIG. 8). This beam is
used to establish the ratio between reference power integral
PI.sub.ref and the known CSA of the reference beam at that depth,
CSA.sub.ref. The same ratio can be applied to the power measured by
the broader measurement beam, PI.sub.meas, in order to determine
the CSA of flow within the beam, provided the measurement beam
encompasses the entire cross-sectional area of the vena contracta.
That is 8 PI meas Flow CSA = PI ref CSA ref ( 8 )
[0080] For example, if the power is four times as great as that
returning from the reference signal (the narrow beam obtained with
the full transducer aperture), the flow CSA at the vena contracta
is four times the known CSA of the reference beam. Since the area
of the reference beam is known, the calibrated power measurement
provides an estimate of the flow CSA at the vena contracta.
[0081] For the illustrative embodiment, the technique used to widen
the beam also reduced the transmit power and receive sensitivity
due to the reduced aperture size. That is, the power measured by
the broader beam is less than that of the reference beam for the
same cross-sectional area of flow, assuming the cross-sectional
area of flow is contained within both the narrow reference and
broader measurement beam (Appendix B). Hence, there is also the
need to correct for the reduction in transmit power and receive
sensitivity simply resulting from the reduction in the aperture of
the transducer. The relationship of the ratios of power to
cross-sectional area for this condition is, 9 CF PI ref Flow CSA =
PI ref CSA ref ( 9 )
[0082] where CF denotes the correction factor accounting for
aperture reduction. This correction factor was determined in vitro
by comparing the backscattered power for both beams for flow
through an orifice smaller in cross-sectional area than both
beam-widths (0.07 cm.sup.2; Appendix B). In this case, as the
transducer aperture was reduced to 11 mm in diameter, the Doppler
power measurement decreased by a factor of 5.0, so that the power
measurement must be multiplied by a correction factor of 5.0 before
it can be compared with the power of the reference beam to
determine the cross-sectional area of the jet at the vena
contracts.
[0083] In order to determine the ratio between power and area, the
backscattered power from the narrow reference beam alternatively
can be captured in the filling phase of the mitral valve when the
valve is wide open (FIG. 1B); that is because the ratio only
requires backscattered power from the flow region of interest
independent from velocity. This potentially would ease the
application of the PVI approach because placing the reference beam
into a small lesion can be difficult. Generally, this indicates
that the power measurement with the reference beam in laminar flow
can be performed independent of velocity and direction of flow.
[0084] The calibration procedure may be simplified by having the
two required beams, the narrow reference and the broad measurement
beam, generated simultaneously. This is done by taking advantage of
a modem digital ultrasound system's capability, the basic principle
being as follows: On transmit, the transducer generates a broad and
relatively uniform beam that sonifies the entire area of flow as
discussed in Appendix D. On receive, the two required beams, the
narrow reference and the broad measurement beam, are generated
simultaneously by the connection of the transducer elements to two
independent digital beam-forming processors. Thus, the Doppler
signal information can be acquired and calibrated in a single
cardiac phase, thereby eliminating the need to perform a separate
calibration operation.
[0085] If a measurement of absolute flow is not required,
calibration is not necessary. For example, the regurgitant
fraction, which is the ratio of regurgitant flow volume to forward
flow volume through the valve, can be measured using this
non-invasive approach. Measurements are obtained of the regurgitant
flow volume and the forward flow volume associated with the valve.
Since the measurement of regurgitant fraction consists of the ratio
of the two measurements, calibration is not necessary.
[0086] System Settings
[0087] In all experimental results to now be described, a 2.5 MHZ
linear array transducer of 96 elements operating at a Doppler
frequency of 1.8 MHZ connected to a Hewlett-Packard 5500 phased
array sector scanner was used. The highest-velocity and
narrowest-spectrum velocity profile just beyond the regurgitant
orifice, from laminar flow at the vena contracta, was recorded with
high pulse repetition frequency (high-PRF) Doppler mode with a
maximum velocity scale of up to 800 cm/s (for details about
high-PRF Doppler see Appendix A). A relatively thin sample volume
of axial gate length of 0.35 cm was used and, to prevent signal
suppression at low flow rates, a low velocity filter (wall filter)
cut-off of 200 Hz (.about.8 cm/s) was selected. Settings of
compress and reject, transmit power, receive gain, depth, and
velocity range were kept constant in the entire in vitro series as
well as within each patient and animal to allow comparison of
returned signals during each series.
[0088] Data Recording and Analysis
[0089] Doppler spectra were stored digitally, converted off-line to
ASCII format and analyzed with MATLAB software (Version 5.1,
MathWorks, Natick, Mass.). Power and power times velocity were then
integrated over all velocities in the velocity spectrum
corresponding to laminar flow in the vena contracta (FIG. 9). In
steady flow, PI (power integral) and PVI (power-velocity integral)
were computed over each vertical line of the spectral Doppler
display (corresponding to a 128-point FFT) and averaging as many as
300 lines in the uncompressed image matrix of power values. For the
experiments, each vertical image line represented a time interval
of 4.9 ms. In pulsatile flow, the PVI was integrated over time to
obtain PVTI (power-velocity time integral) as a measure of
regurgitant blood flow volume within one cardiac phase. The
difference between PI and PVI is seen in FIG. 9: PI is relatively
constant during flow, given the constant in vitro orifice area,
whereas PVI, reflecting the flow rate, varies over time and mirrors
the velocity as expected for a fixed orifice.
[0090] Results
[0091] In vitro and in vivo test results for the Doppler
power-velocity based technique described above, preceded by a brief
explanation of the experimental setup, are presented in the
following sections.
[0092] In Vitro Setup
[0093] For in vitro validation of the Doppler power-velocity
concept for quantification of flow area, flow rate and flow volume
for a dynamic orifice, a flow phantom was built to simulate the
condition of regurgitant flow at the mitral valve. The flow phantom
(FIG. 10) consisted of a variable diameter orifice located 10 cm
from the ultrasound transducer acoustic window (typical distance in
vivo of a mitral regurgitant orifice from the transducer placed on
the chest surface; see FIG. 2A), with flow passing from a large
cylindrical chamber (5.7 cm diameter; mimicking the LV) to an
unconfined receiving chamber (mimicking the LA). Circular orifice
areas of 0.2, 0.5 and 0.8 cm.sup.2 were studied at 5 steady flow
rates from 20 to 60 ml/s produced by a piston pump (modified Mark
IV Powerinjector, Medrad, Inc., Indianola, Pa.) that minimized
cavitation, as opposed to impeller-based pumps which were found to
produce microairbubbles by cavitation, and with 6 parabolic pulses
of regurgitant stroke volumes of 20 to 70 ml (4 different-shaped
flow profiles for each stroke volume) produced by syringe
injections. Initial studies with outdated blood or blood analogs
containing surfactant produced microairbubbles due to cavitation as
fluid encountered the pressure drop at the orifice. These
microairbubbles severely contaminated the power measurements due to
their strong backscattering properties. This was resolved using
degassed distilled water with 19% glycerol (resulting specific
gravity=1.043 gm/cm.sup.3) with a backscattering coefficient
equivalent to that of whole human blood, produced by adding 48,600
polystyrene microspheres/ml (25.2 micron diameter; catalog no.
DC-25; Duke Scientific Corp., Palo Alto, Calif.).
[0094] This flow phantom consisting of two chambers separated by an
orifice models the similar condition of valvular stenosis and
trans-septal shunting, and hence the results are equally applicable
to the application of this procedure to quantification of stenotic
valve opening area and shunt area and flow.
[0095] In Vitro Results
[0096] For all orifice sizes and flow rates, a clear narrow bright
band of intensities at the highest velocities corresponding to the
laminar flow at the vena contracta was detected in both steady and
pulsatile flow. FIG. 12A shows that, as proposed, there was a
linear proportionality between the power integral and the
regurgitant orifice area (ROA) up to and including 0.5 cm.sup.2 in
area (r=0.99); an extreme orifice area of 0.8 cm.sup.2 was
incompletely measured with the beam size used in the experiments
(PI below the regression line for the other orifices). This,
however, illustrates the importance of having a measurement beam
that is broad enough to fully ensonify the area of the vena
contracta.
[0097] These power values were measured using the expanded beam
created by reducing transducer aperture to a diameter of 7 mm.
Calibration was achieved with the narrow reference beam produced by
the full aperture. FIG. 12B demonstrates that this narrow beam fit
within all the orifices used and therefore returned the same power
regardless of orifice area and flow rate over the 32 combinations
studied. This narrow beam was then used to create the following
calibrations:
ROA=c.sub.cal.cndot.measurement beam PI
Q=c.sub.cal.cndot.measurement beam PVI
RSV=c.sub.cal.cndot.measurement beam PVTI
[0098] where c.sub.cal=(reference beam CSA/reference beam PI) times
the correction factor (CF) needed to account for the decrease in
power related to aperture reduction (see Eq. 9). The correction
factor CF and the reference beam CSA are determined in vitro; only
reference and measurement beam PI need to be measured in vivo. In
vitro, reference beam CSA was 0.125 cm.sup.2 and CF was 63
(Appendix B, 7 mm aperture). With calibration, PI values were
converted to regurgitant orifice area (ROA) and correlated and
agreed well to actual ROA less than 0.8 cm.sup.2 (FIG. 12A,
right-hand axis, y=0.84x+0.01 cm.sup.2, SEE=0.01 cm.sup.2), with
slight underestimation of the anatomic orifice areas of 0.2 and 0.5
cm.sup.2 since the cross-sectional area of the vena contracta is
known to be slightly smaller as determined by a contraction
coefficient.
[0099] For the 0.07, 0.2 and 0.5 cm.sup.2 orifices, flow rates
calculated from PVI correlated linearly with actual flow rates
(FIG.13; r=0.99, y=0.95x+0.21 ml/s, SEE=2.5 ml/s). Similar
correlation (FIG. 14; r=0.99, y=0.92x+1.3 ml, SEE=2.6 ml) was
observed for the 72 studied pulsatile flow volumes calculated from
PVTI.
[0100] With increasing velocity, a broader velocity spectrum was
observed, raising the possibility of turbulent flow. However, as
shown in Appendix C, the spectral bandwidth, measured as the
central range of velocities in which 95% of scatterers travel (95%
of power), was always roughly 20% of the power-weighted mean
velocity (PWMV), indicating effectively laminar flow.
[0101] In Vivo Experiments
[0102] In vivo results are based both on animal experiments and on
human patient studies. The animal experiments are presented first.
The patient results, including data derived from three-dimensional
(3D) ultrasound and magnetic resonance imaging (MRI) techniques for
independent reference follow.
[0103] Animal Study Setup
[0104] In order to retain the native leaflets of the mitral valve,
most comparable to the physiologic situation, and to measure MR
volume directly, we developed the new canine model (FIG. 15). A
nondistensible left atrial chamber was sutured to the native mitral
annulus via a Dacron sewing ring, and the atrial walls sutured
tightly around it to prevent leaking. This chamber was attached via
nondistensible tubing to a vertical 1.0 cm diameter column. In the
presence of a rigid left atrium, the entire MR stroke volume (MRSV)
produced a vertical fluid excursion in the attached column with
each systole (1.3 cm=1 ml), this excursion was videotaped and
measured. Using this system, a total of 36 different hemodynamic
stages were analyzed in 3 dogs, in each of which a regurgitant
orifice of different size (0.12 to 0.21 cm.sup.2) was cut into the
anterior mitral valve leaflet and afterload changed by clamping the
aorta to vary the amount of regurgitant flow.
[0105] All procedures were approved by the institutional Animal
Care Committee based on NIH (National Institute of Health)
principles.
[0106] Animal Study Results
[0107] During each hemodynamic stage, regurgitant stroke volume as
measured based on the video camera recordings was stable over the
time of recording (10-20 cycle). FIG. 16 shows results for MRSV
calculated from the power measurements versus directly measured
values of MRSV (left atrial column excursions) for 36 hemodynamic
stages in all three animals. Actual values of mitral regurgitation
stroke volume ranged from 4 to 21 ml. Calculated and actual values
correlated excellent (y=0.98x+0.28 ml, r=0.98, SEE=0.89 ml).
[0108] Patient Study Setup
[0109] To demonstrate that the Doppler power principle can in fact
be applied in patients, the amount of mitral regurgitant flow
volume (MRSV) that occurs in each systolic phase of the, heart
cycle was determined in eleven patients (age 65.+-.13 years; 8
male, 3 female) with at least mild to moderate mitral regurgitation
of different etiologies, using the same methods applied in vitro.
In all patients, values for MRSV by Doppler power were compared
with those calculated from left ventricular ejection volume by
three-dimensional echocardiography minus forward aortic stroke
volume derived from left ventricular outflow tract CSA times the
time integral of the power-weighted mean velocity measured in the
LV outflow tract, which takes into account the velocity spectrum
within the beam. PVI measurements were also compared with estimates
of mitral regurgitant stroke volume obtained using MRI in three
patients. To date the MRI approach represents the most trusted
procedure for quantification of regurgitant flow. However, it is
extremely time consuming, requiring patient confinement within the
bore of the magnet of 1 hour, and is prohibitively expensive and
not practical for this application.
[0110] Patient Study Results
[0111] In all patients, a satisfactory high-PRF Doppler signal with
a narrow velocity spectrum from laminar flow at the vena contracta
was obtained; including those with eccentrically directed jets.
FIG. 17 shows an example of such a velocity spectrum in a patient
with functional mitral regurgitation. Of note is that the power
integral has early and late systolic peaks and mid-systolic
decrease, which is the recognized pattern for regurgitant orifice
area in such patients since the two valve leaflets are pushed
together in mid-systole by the peak LV pressure, which decreases
the orifice area.
[0112] Values of mitral regurgitant stroke volume in the eleven
patients as measured by the reference techniques ranged from 8.0 to
50.1 ml representing mild to severe regurgitation. Correlation
between the PVTI estimates of MRSV and reference values are shown
in FIG. 18 (r=0.99, y=0.72x+3.4 ml, SEE=1.45 ml). The correlation
is good for flow volumes up to 20 ml. However, there is
underestimation of regurgitant volumes above 20 ml that is
attributed to the limited measurement beam size used in these
experiments; that is, the larger regurgitant flow volumes
correspond to large regurgitant orifices which exceeded the
cross-sectional area of the measurement beam. It is anticipated
that with a broader measurement beam this underestimation of the
higher regurgitant stroke volumes can be eliminated.
CONCLUSIONS
[0113] Ultrasound over the past twenty years has proven to be
extremely valuable as a noninvasive medical diagnostic tool not
only for cardio-vascular diseases but also for prenatal and
gynecologic examinations, and for diseases of the liver and kidneys
among others. Some new uses include tissue characterization, breast
cancer identification, and eye examinations. However, though the
target of many researchers worldwide, the direct and noninvasive
quantification of regurgitant pathologic volume flow in the human
heart is still not possible. Existing ultrasound approaches, either
based on simplification of the regurgitant pathologic flow model or
on complex multi-step procedures, do not measure the flow directly
at the lesion of the valve and have failed to provide satisfactory
results. For example, direct measurement of the regurgitant flow,
is dependent upon the knowledge of the area of flow, which is
difficult to measure due to its complex and dynamically changing
geometry; until the development of this invention no ultrasound
technique has been available to provide such a measurement in a
reliable and useful manner.
[0114] The innovative approach of this invention is based on the
integral of backscattered acoustic power, which, in principle,
provides a measure of flow cross-sectional area. Although the basic
physics of flow and ultrasound indicate the integral of Doppler
power times velocity can measure flow rate and flow volume,
investigators have felt this principle could not be applied to
regurgitant pathologic jets comprised of high-velocity flow in the
human heart because backscattered power is variably increased by
turbulence and entrainment of fluid. For that reason previous
studies using Doppler power evaluated only flows that were
restricted to a conduit (tube, large vessel, or ventricular outflow
tract) and therefore considered laminar. This invention
demonstrates that this limitation can be overcome by integrating
power times velocity in the narrow velocity spectrum obtained from
flow passing through a thin sample volume placed in the vena
contracta, that is, at the origin of the regurgitant jet, where
flow is mostly laminar since turbulence has not yet developed.
[0115] The approach of this invention has several advantages.
First, regurgitant flow rate and volume can be measured using only
the backscattered power and velocity information as obtained by
ultrasound Doppler. Because, integrating backscattered power times
velocity over the Doppler spectrum automatically contains a measure
of the flow cross-sectional area due to the proportionality of
backscattered power to flow area, this approach overcomes the
limitations of most techniques that require a separate area
measurement. Second, the thin Doppler sample volume can be placed
directly at the lesion where regurgitant flow occurs, making
possible the most direct and therefore accurate measurement of
regurgitant flow area and flow volume. Third, with this approach
there are no simplifying assumptions about the velocity
distribution across the flow area, because the Doppler beam
effectively integrates all incremental areas of different flow
rates. There are also no concerns about the spectrum of velocities
because the approach integrates backscattered power times velocity
over the entire velocity spectrum, as long as the spectrum is
reasonably narrow indicating mostly laminar flow. Finally, the
power-velocity measurement, is independent from variations in the
angle .theta. between the direction of flow and the main axis of
the Doppler beam. It has been shown that this principle is correct
at least for angles from 0 to 45.degree..
[0116] It is important to realize that the size and shape of a
regurgitant orifice in vivo is varying throughout the period of
regurgitation, thereby limiting the ability of existing techniques
which rely upon a single estimate of the orifice area, since no
satisfactory procedures were available to record these dynamic
changes. The beauty of the PVI approach is that, because the
Doppler beam effectively integrates all incremental areas of
different flow rates independent of their local distribution across
the beam area, there are no longer concerns about the complex shape
and size of the dynamic flow cross-sectional area or the
regurgitant orifice. This is also important because different
shapes and sizes of regurgitant orifices caused different complex
flow fields around the valve lesion thereby limiting the accuracy
of other techniques based on rather simplistic geometric flow field
assumptions.
[0117] This concept was applied in an ideal in vitro environment,
in a more realistic experimental in vivo setting with a direct
goldstandard, as well as in patients, with very good results,
demonstrating a linear proportionality between flow cross-sectional
area and the Doppler power integral (PI), between regurgitant flow
rate and the integral of Doppler power times velocity (PVI), and
between regurgitant stroke volume and the time integral of the
integral of Doppler power times velocity (PVTI), all measured
directly at the vena contracta. These results, and the narrow
velocity spectra observed even for relatively small orifices and
eccentric jets in patients, suggest that no significant turbulence
develops at the vena contracta that would interfere with this
proportionality. Besides demonstrating the proof of the PVI
concept, the practicability and accuracy of this new approach was
successfully demonstrated in patients with various mitral valve
diseases leading to mitral regurgitation.
[0118] Future systems may be developed which simplify the procedure
for routine clinical application. The described system has an
ultrasound transducer capable of creating a broad and uniform
measurement beam and a narrow reference beam for calibration
purposes. Thereafter a manual, partially automated or fully
automated analysis of the Doppler signal is performed by the
system, identifying the narrow Doppler spectrum associated with
laminar flow in the vena contracta and calculating PI, PVI, and
PVTI. Finally, the results of the measurement beam are calibrated
taking into account transmit power, receive gain, and other
relevant system settings, thereby producing and displaying
measurements of diagnostic importance such as peak cross-sectional
area of flow, mean cross-sectional area of flow, peak flow rate,
mean flow rate, stroke volume and regurgitant fraction.
[0119] Beyond the presented application in mitral valve
regurgitation, this new concept can be applied to other diseases
associated with flow through dynamic orifices; for example
regurgitant flow of each of the four heart valves, as well as
valvular stenosis, septal defects with shunt flow, and peripheral
vascular disease with vessel obstruction, where valvular stenosis
refers to restricted forward flow due to a condition that prevents
the valve from opening completely and a septal defect refers to a
pathologic orifice in the septum that separates the right and the
left side of the heart. In particular, accurate noninvasive
measurement of stenotic (restricted) valve opening areas based on
the calibrated power integral would critically improve diagnosis
and treatment of stenotic valve disease. Ultimately, this PVI
approach could be used to obtain regurgitant fraction as the ratio
of regurgitant flow volume to normal forward flow volume; this
would require a beam encompassing the cross-sectional area of
forward flow when the valve is open. The beauty of this approach
would be that both regurgitant and forward volume could be
determined from the same sample volume. By adding regurgitant and
forward flow volume at the mitral valve, cardiac output, one of the
most important clinical parameters, is obtained, where cardiac
output refers to the volume of blood pumped by the heart in a
cardiac cycle.
[0120] Thus, the PVI concept can overcome the limitations of
existing techniques that are invasive, costly, and inaccurate and
has the potential to replace them.
* * * * *