Note: Descriptions are shown in the official language in which they were submitted.
CA 02461972 2004-03-23
-1-
Weighted Gradient Method and Svstem for Dias~nosins~ Disease
Cross References to Related Aaulications
This application claims priority from Canadian application serial no.
2,423,442 filed March 26, 2003, the contents of which are incorporated herein
by reference.
Field of the invention
This invention relates to a method for detecting and diagnosing disease
states in living organisms and specifically relates to diagnosis of disease by
measuring electrical properties of body parts.
Backs~round of the invention
Several methods exist for diagnosing disease that involve measuring a
physical property of a part of the body. A change in such a physical property
can signal the presence of disease. For example, x-ray techniques measure
tissue physical density, ultrasound measures acoustic density, and thermal
sensing techniques measures differences in tissue heat generation and
conduction. Other properties are electrical, such as the impedance of a body
part that is related to the resistance that the body part offers to the flow
of
electrical current through it.
CA 02461972 2004-03-23
-2-
Values of electrical impedance of various body tissues are well known
through studies on intact humans or from excised tissue made available
following therapeutic surgical procedures. In addition, it is well documented
that a decrease in electrical impedance occurs in tissue as it undergoes
cancerous changes. This finding is consistent over many animal species and
tissue types, including, for example human breast cancers.
There have been a number of reports of attempts to detect breast
tumors using electrical impedance imaging, such as, for example, U.S. Pat.
No. 4,486,835. However, there are basic problems when trying to construct an
image from impedance data. Electric current does not proceed in straight lines
or in a single plane; it follows the path of least resistance, which is
inevitably
irregular and three-dimensional. As a result, the mathematics for constructing
the impedance is very complex and requires simplifying assumptions that
greatly decrease image fidelity and resolution.
Despite such difficulties, a method that permits comparisons of
electrical properties for diagnostic purposes has been developed that involves
homologous body parts, i.e., body parts that are substantially similar, such
as
a left breast and a right breast. In this method, the impedance of a body part
of a patient is compared to the impedance of the homologous body part of the
same patient. One technique for screening and diagnosing diseased states
within the body using electrical impedance is disclosed in U.S. Pat. No.
CA 02461972 2004-03-23
-3-
6,122,544, which is incorporated herein by reference. In this patent, data are
obtained from two anatomically homologous body regions, one of which may
be affected by disease. Differences in the electrical properties of the two
homologous body parts could signal disease. One subset of the data so
obtained is processed and analyzed by structuring the data values as
elements of an n x n impedance matrix. The matrices can be further
characterized by their eigenvalues and eigenvectors. These matrices and/or
their eigenvalues and eigenvectors can be subjected to a pattern recognition
process to match for known normal or disease matrix or eigenvalue and
eigenvectors patterns. The matrices and/or their eigenvalues and
eigenvectors derived from each homologous body region can also be
compared, respectively, to each other using various analytical methods and
then subjected to criteria established for differentiating normal from
diseased
states.
Published international patent application, PCT/CA01/01788, which is
incorporated herein by reference, discloses a breast electrode array for
diagnosing the presence of a disease state in a living organism, wherein the
electrode array comprises a flexible body, a plurality of flexible arms
extending from the body, and a plurality of electrodes provided by the
plurality
of flexible arms, wherein the electrodes are arranged on the arms to obtain
impedance measurements between respective electrodes. In one
embodiment, the plurality of flexible arms are spaced around the flexible body
and are provided with an electrode pair. In operation, the electrodes are
CA 02461972 2004-03-23
-4-
selected so that the impedance data obtained will include elements of an
n x n impedance matrix, plus other impedance values that are typically
obtained with tetrapolar impedance measurements. Tetrapolar impedance
measurements are associated with injecting current between so called current
electrodes and measuring a voltage drop between associated electrodes. In
a preferred embodiment, the differences between corresponding homologous
impedance measurements in the two body parts are compared in a variety of
ways that allow the calculation of metrics that can serve to either indicate
the
presence of disease or localize the disease to a specific breast quadrant or
sector. The impedance differences are also displayed graphically, for example
in a frontal plane representation of the breast by partitioning the impedance
differences into pixel elements throughout the plane.
Despite the attractive features of this method of diagnosing disease in
one of a homologous pair of body parts, there are some problems associated
with this straightforward implementation. In particular, the current path
through the body part, whether healthy or not, as the current flows from one
electrode to the other is, in general, complex. It encompasses to a certain
extent, all areas of the body part. In the aforementioned method, this
complexity is addressed by simplifying assumptions. This simplification may
affect the ability of the method to detect the disease.
CA 02461972 2004-03-23
-5-
Summary of the invention
The present invention is directed to an improved method for detecting
and diagnosing disease states in a living organism by using a set of
electrical
impedance measurements. The method is based on the realistic distribution
of electric current in the body part. For each impedance measurement, the
approximate current distribution is obtained by a numerical computation using
a representation of a body part structure, or by the direct measurement
performed on a physical model or a control subject's body part. This obtained
current distribution is further used to correlate impedances obtained by
direct
measurements to different areas in the body part.
To achieve this goal, the subject body part is subdivided into a number
of small regions called finite elements. For each of the elements and for each
of the electrode pairs used to inject current into the body part, a weight
factor
(obtained by computing or measuring the current density in the element),
reflecting the position of the element within the body part, is calculated and
stored. Each element has one weight factor for each current injection. Larger
weight factors are associated with current injections that result in larger
current densities in a particular element. Thus, current injecting scenarios
associated with larger weights at a particular element are given greater
consideration when detecting disease. The weights are typically calculated or
measured with the assumption that there is no disease present. At the same
time, baseline impedances associated with each of the current injections are
obtained. The weights and baseline impedances for each of the current
CA 02461972 2004-03-23
injection scenarios are stored in the database and used when a diagnosis is
made following the measurement of the actual impedances of the subject's
body part. For each element, the diagnostic is the sum over all current
injections of weight multiplied by the ratio of baseline to measured
impedance.
This sum is referred to as a Weighted Element Value (WEVaI). The higher
the value of the sum is, the higher is the probability of the disease at the
location of a particular element. Elements are grouped according to known
physical characteristics and a sum for each of the groups is obtained.
Comparing sums of homologous regions may point to a presence of disease
in the body part.
In particular, a system and method for diagnosing the possibility of
disease in a body part is described herein. The system includes an electrode
array by which an electrical property of the body part may be measured, such
as a measured impedance. The system further includes a grid module for
representing the body part with a grid having a plurality of finite elements,
and
for obtaining a baseline electrical property using a model of the body part,
such as a baseline impedance. The system also includes a weight module for
using the model of the body part to compute a set of weights associated with
a particular one of the plurality of finite elements, each weight in the set
derived from a particular current injection electrode pair selection. A
diagnostic module computes a diagnostic at the particular finite element to
diagnose the possibility of disease in the body part, the diagnostic being a
CA 02461972 2004-03-23
_ 7 _
function of the measured electrical property, the baseline electrical property
and the set of weights.
Brief description of the drawinos
Figure 1A shows the components of a basic tetrapolar measurement;
Figure 1 B is a block diagram of a system for detecting and diagnosing
disease in a body part;
Figure 1 C is a block data flow diagram of a method for detecting and
diagnosing disease in a body part;
Figure 2 is a sample finite element grid produced by the grid module of
Figure 1 B, the grid representing a body part that can be used to calculate
baseline electrical properties;
Figure 3 is a block data flow diagram of the grid module of Figure 1 B,
in one embodiment of the present invention that employs a numerical finite
element method;
Figure 4 is a block data flow diagram of the diagnostic module of
Figure 1 B, in one embodiment of the present invention;
CA 02461972 2004-03-23
8
Figure 5 is a flowchart illustrating the method steps performed by the
diagnostic system of Figure 1 B to diagnose disease; and
Figures 6A and 6B are sample WEVaI plots for an actual subject that
can be used to detect breast cancer.
Detailed description of the invention
Figure 1A shows a schematic of components used to perform a
tetrapolar impedance measurement, which measurements are used for
detecting and diagnosing disease, as described in more detail below. Figures
1 B and 1 C show a block diagram of a system 10 and an outline of a method
for detecting and diagnosing disease in a body part, such as breast cancer.
The method uses impedance measurements taken from a multi-channel
impedance measuring instrument 11 with a pair of electrode arrays 12, like
the one described in PCTICA01/01788, a grid module 14 and a diagnostic
module 16.
Referring to Figure 1A, a single electrical impedance measurement is
performed using four electrodes. One pair of electrodes 1 is used for the
application of current I, and the other pair of electrodes 2 is used to
measure
the voltage V that is produced across a material, such as breast tissue 3, by
the current. The current I flowing between electrodes 1 is indicated by the
arrows 4. The impedance Z is the ratio of V to I; i.e., Z = V/1. By using
CA 02461972 2004-03-23
_g_
separate electrode pairs for current injection and voltage measurement,
polarization effects at the voltage measurement electrodes 2 are minimized
and a more accurate measurement of impedance can be produced. It should
be understood that, in general, the voltage electrodes 2 need not be disposed
between the two current electrodes 1.
Impedance consists of two components, resistance and capacitive
reactance (or equivalently, the magnitude of impedance and its phase angle).
Both components are measured and analyzed in the present invention.
However, in examples described below, only resistance is used and
interchangeably referred to as either resistance or the more general term
impedance.
As has been noted above, by performing tetrapolar measurements in
which separate electrode pairs are used for current injection and voltage
measurement, polarization effects at the voltage measurement electrodes 2
are minimized and more accurate measurements of impedance can be
performed. However, there may be some embodiments in which bipolar,
instead of a tetrapolar, measurements can be performed as part of the
general method for diagnosing disease discussed below. If bipolar
measurements are performed, a correction factor can be used that corrects
for the polarization effects arising from skin-to-electrode interface.
CA 02461972 2004-03-23
-10-
Figure 1 B shows a schematic of the electrode array 12. Eight current
injection electrodes 13, and eight associated voltage measurement electrodes
15 are shown. In general, there are ne current injection electrodes and ne
associated voltage measurement electrodes in the electrode array. The
electrodes are applied on the body part, each of the current injection
electrodes being associated with the adjacent voltage measurement
electrode. Impedance is measured between two voltage electrodes when the
current is injected between associated current electrodes. Since there are
n~, = ne~(ne-1 )l2 pairs of current injection electrodes, and an equal number
of
voltage measurement electrode pairs, the total number of independent current
injections and related impedances is n~,. It should be understood that the
electrode array shown is but one possible electrode array. Other electrode
arrays may also be used.
As discussed in more detail below, the grid module 14 uses a
numerical or physical model of a baseline (idealized or reference) body part
to
compute baseline values. In particular, at step (66), baseline impedances and
associated gradients for the baseline body part are calculated in the grid
module 14. As detailed below, the associated gradients can be used to
calculate current densities at each finite element. The baseline impedances
for each of the n~, current injections, and the associated current densities
for
each of the finite elements and for each of the nc, current injections are
stored
in a baseline body parts database 17.
CA 02461972 2004-03-23
-11-
At step (68), the impedance is measured n~~ times resulting in the set
of values, f Z;' , ZZ , ... , Z,G }, where ZM is the impedance measured
between the voltage electrodes associated with the ~~" current injection
electrode pair when current is injected between that current injection
electrode
pair, as required in tetrapolar impedance measurement.
The grid module 14 includes software and/or hardware for representing
the body part with a grid of elements that are so small that the voltage
gradient during arbitrary current injection is approximately constant within
any
single element. For example, if the body part is modeled as a two-
dimensional surface, then the grid can be composed of triangles that "tile"
the
surface. Alternatively, the body part can be modeled by a three-dimensional
grid whose elements are tetrahedrons, for example. Each finite element is
associated with a plurality of nodes, typically on the perimeter of the finite
element. As well, each finite element is characterized by its electrical
material
property, namely resistivity and/or permittivity. Adjacent elements share the
nodes associated with the common side or face. When the elements are
small enough to ensure that the current density throughout the element is
constant for each of the current injections, the voltage gradient throughout
the
element is also constant and proportional to the current density.
The grid module 14 also includes software and/or hardware for deriving
the current density for each of the elements in the grid. It does this by
CA 02461972 2004-03-23
-12-
calculating the current density using a numerical or physical model, or by
using population study information, as discussed in more detail below.
The diagnostic module 16 includes software and/or hardware for
detecting the presence of a tumor in the body part at step (70). As described
in more detail below, the diagnosis is based on a diagnostic that is a
function
of the impedance measurements obtained from a subject using the
impedance measuring instrument 11, and a weighting factor derived from the
estimated value of the current density throughout the body part, obtained
using grid module 14.
Figure 2 shows a representation of the baseline body part divided into
a grid 80 composed of a plurality of finite elements 82. Once the body part is
subdivided using grid module 14 into a number of finite elements 82, there are
several methods that can be used to calculate baseline values, such as the
current density associated with a particular current injection and with a
particular finite element 82 of the grid 80. Figure 2 shows one embodiment of
the present invention in which several thousand finite elements 82 are used,
as required to justify linearizing the equations used to numerically compute
the relevant electrical properties.
The preferred method used by the grid module 14 to associate a
voltage gradient with a particular finite element 82 is a numerical finite
element method that assumes that the resistivity of the body part is uniform.
CA 02461972 2004-03-23
-13-
The method numerically solves Laplace's equation, known to those of
ordinary skill, to compute the electric potential at the nodes of the finite
element grid from which the electric voltage gradient can be obtained. Due to
uniform resistivity, current density is proportional to the voltage gradient
everywhere in the body part.
A second method that can be used by the grid module 14 is related to
the last method, except that instead of assuming a uniform resistivity, more
realistic resistivities andlor permittivities can be used that reflect the
known
internal structure of the body part. In this case the current density is
proportional to the electric voltage gradient in each of the elements, but the
voltage gradient to current density ratio depends on the resistivity and/or
reactivity associated with the particular finite element 82.
The third method involves using a physical model of a typical breast.
This typical breast acts as a baseline representation of the body part. The
model is designed so that the measured impedance matrix is close to the
average impedance matrix for the normal subject with the body part of the
particular size. Each finite element 82 obtained using the grid module 14 is
associated with the particular location (x, y and z coordinates) in the
physical
model. The current density at each of the finite elements 82 and for each of
the current injections is obtained using one of the available instruments for
measuring the current density. The current density instrument, for example,
can be combined with magnetic resonance imaging (MRI) to measure and
CA 02461972 2004-03-23
, i
-14-
display the current density superimposed on the MRI image at any location of
the body part model.
The fourth method is similar to the third method except that the
measurement of the current density for each current injection and at the
location of each of the finite elements 82 defined by the grid module 14 is
performed on the body part of an actual control subject. For example, the
same combination of instruments as above can be used to measure and
display the current density' superimposed on the MRI image at any location in
the actual body part.
Figure 3 shows a block data flow diagram of the grid module 14 in the
preferred embodiment of the invention where it includes a finite element
analysis module 28 and a gradient module 30.
In the preferred embodiment of the invention, for any single current
injection, a finite element method is used to estimate baseline values for
electric potential gradients and resulting current densities in each of the
elements. In addition, the grid module 14 uses the finite element method to
compute the baseline impedance. More generally, the baseline impedance
refers to the impedance calculated by the grid module 14 (denoted by Z~, for
the jt" electrode pair) using an appropriate physical or numerical model, as
CA 02461972 2004-03-23
-15-
distinguished from the measured impedance, Z"', obtained by a
measurement on a subject using an electrode array.
The finite element analysis module 28 includes hardware and/or
software that employs various boundary conditions, corresponding to the
injections of current between the various pairs of current injection
electrodes
13 (Figure 1 B), to compute the electric potential at all the nodes in the
grid.
The node voltage V~; is the voltage that arises at the node j when a current
injection i is applied, where the it" current injection refers to the
injection of
current between the i~' current injection electrode pair.
Specifically, the finite element analysis module 28 includes a finite
element grid generator 29, a boundary conditions generator 31 and a finite
element equation solver 33. The finite element grid generator 29 generates a
grid 80 of finite elements 82 that spans a representation of the body part.
Position on the representation of the body part can be discretized if each
finite
element is associated with several nodes, typically on the perimeter of the
finite element.
To compute the potential, V, as a function of position on the grid,
Laplace's equation o2V = 0 is solved using a numerical finite element method.
The boundary conditions generator 31 assigns boundary conditions
CA 02461972 2004-03-23
-16-
corresponding to the various n~, current injections. The finite element
equation solver 33 employs the numerical finite element method for solving
Laplace's equation. Many different types of such methods can be used, such
as a Lax differencing scheme for solving partial differential equations.
Several
other techniques known to those of ordinary skill in the art can be utilized.
In addition to finding the electric potential as a function of node
position, the grid module 14 also finds voltage differences between voltage
measurement electrodes 15. In particular, using boundary conditions
corresponding to the current injected by the first pair of current injection
electrodes yields V,, the voltage drops between the first pair of voltage
measurement electrodes. Using boundary conditions corresponding to the
current injected by the second pair of current injection electrodes yields VZ,
the voltage drop between the second pair of voltage measurement electrodes.
Continuing in this manner yields all n~, voltages {Y, Vz, ... , Vn~}. Each
time
Laplace's equation is solved, the finite element method yields the potential
at
every node of the grid as well. The node voltage V~; is the voltage that
arises
at the node j when a current injection i is applied. The gradient module 30
utilizes the calculated node voltages to find an estimated current density at
the element k for the current injection i, J;k. The grid module 14 similarly
obtains all nc, impedances {Z, , ZZ , ... , Zn~ } and all the current
densities
{J,k , JZk , ..., Jn~k}, at the finite element k. In particular, to obtain
J;k, where
CA 02461972 2004-03-23
-17-
J;k is the magnitude of the current density in the 1~" finite element for the
current injection i, the gradient module 30 uses the electric potential at
each
node associated with finite element k. To this end, the magnitude of the
gradient of the electric potential, which is equal to the magnitude of the
electric field, is first obtained by a voltage gradient calculator 37.
For example, supposing the element to be two dimensional with
potential V=~(x,y), then E=ID~ where E is the magnitude of the electric
field. The voltage gradient calculator 37 can obtain E as follows. In the
(x,y,V) coordinate system, if 8 is the angle between k, the unit normal in the
V direction, and the perpendicular to the surface Y =q~(x, y) , then ~e = ID$I
.
To see this, an auxiliary function F(x,y,Y)=Y-~(x,y) can be introduced.
The quantity OF / I DF I is a normal vector perpendicular to the level surface
F(x,y,Y)=coast., or, with coast. =0, a normal vector perpendicular to the
surface y - ~ (x,y) . Then,
DV
sing k
_ X
IDYI
cos8 k.OY
Ivvl
I2
=E
CA 02461972 2004-03-23
-18-
When employing the finite element analysis, the finite element analysis
module 28 can either assume the body part to have a uniform resistance
and/or reactance, or the resistance and/or reactance can be taken to be non-
uniform to reflect the known structure of the body part.
A current density calculator 35 calculates the magnitude of the current
density J from the magnitude of the electric field E and the tissue
resistivity p
using the microscopic version of Ohm's Law stating that at every point,
J=Elp.
Figure 4 shows a block data flow diagram of the diagnostic module 16
of Fig. 1 B, in one embodiment of the present invention. The diagnostic
module 16 includes a weight module 22, an averaging module 24 and a
comparator 26.
As discussed previously, the diagnostic module 16 computes a
Weighted Element Value (WEVaI) parameter (diagnostic) at each of the finite
elements 82 of the grid 80 representing the body part, and utilizes the
diagnostic to diagnose the possibility of disease in the body part. The
diagnostic is a function of the impedances and current densities calculated
and/or measured for the baseline body part and impedances measured on the
body part of the subject.
CA 02461972 2004-03-23
-19-
The weight module 22 includes software and/or hardware for
calculating weights for the element k and the current injection i, w;~ , given
by
lN~k = n 'I ik .
The quantity Jlk is the magnitude of the current density, which exists at the
finite element k when the reference current is applied between the first pair
of
current injection electrodes. The quantity JZk is the magnitude of the current
density, which exists at the finite element k when the reference current is
applied between the second pair of current injection electrodes, and so on.
The averaging module 24 includes software andlor hardware for
calculating a weighted average of a function ~'(Z1,ZM) . The diagnostic at the
finite element k is defined to be
(J'k~=~,wk.f(Z;~ZM) .
im
The diagnostic ( fk~ is referred to as the Weighted Element Value (WEVaI).
The quantity Z, is the impedance between the first pair of electrodes for the
baseline body part. The quantity Z2 is the impedance between the second
pair of electrodes for the baseline body part, and so on. The Z, can be
obtained using a numerical calculation or using a physical model (an
artificial
reproduction or the real body part of a control subject). The ZM are obtained
by direct measurement on the body part of a subject using an electrode array.
CA 02461972 2004-03-23
-20-
In the preferred embodiment of the present invention, the function f(Z;,Z"')
is
Z.
M _ a
.f (Z,,Za )- M
Z;
It should be understood that other functions ~' might be used in other
embodiments, including functions that are independent of the baseline values
Z;. It should be further understood that the diagnostic module 16 can
condition the raw measurements Z;, such as by standardizing with a factor,
etc, to find the diagnostic. Thus, in one embodiment, the function can be
given by
f (Z;~ZM ) _ ~~
for some appropriate factor, a, used to condition the raw data, which
conditioned data may be used to compute the diagnostic.
In a human subject, some body parts have homology in the body. For
example, in females, the right breast has a homolog, namely the left breast.
In a preferred embodiment of the invention, (fx~ is averaged over all the
finite
elements of the right breast to yield (fri~,~ ~ , and all the finite elements
of the
left breast to yield (f~~ . In a different embodiment, (.fri~,t ~ can refer to
an
average over finite elements belonging to a particular region within the right
breast.
CA 02461972 2004-03-23
-21 -
More generally, if the N finite elements comprising the grid are not all of
equal size, the average is given by
N
~fri~t ~ - ~ pk (fk ~~ where the probabilities pk are given by
m
pk - /1.A (k)Vk /vA .
In this last expression, xA(k) is the characteristic function for a region
A of the body part:
1, if finite element k c A
'~A (k) - 0, otherwise
and Vk and VA are the volumes (if the grid is three dimensional) or the areas
(if the grid is two-dimensional) of finite element k and region A,
respectively.
The measured impedances in the body part are expected to be
somewhat different from the values measured in the homologous body part.
However, these differences are expected to be more pronounced if only one
of these body parts contains a malignant tumor.
The comparator 26 includes hardware and/or software for comparing
(f~~ to (f~~~~ to diagnose the possibility of disease. For example, if breast
cancer is being diagnosed and if it is assumed that at least one breast is non-
CA 02461972 2004-03-23
-22-
cancerous, then a difference between ~f,~~ and (f,;gh~ ) may be due to a
change in the electrical properties of one breast brought about by the
presence of a cancer.
The comparator 26 calculates the absolute difference hf,;~,, )- (fe ft )I or a
relative difference such as ~( f,;~,t ) - ( f,~ )~ 2 ~ ~( fri~,~ ) + ( f,~ )~~
that is indicative
of the possibility of disease in the body part or the homologous body part.
Where there is a significant difference, further analysis can be performed to
discern which of the homologous pairs may be cancerous. For example, as
described above, it is known that the electrical properties of cancerous
tissue
deviate from the norm in a predictable way. Thus, the body part having
electrical properties more like those of a cancerous body part can be suspect.
It should be understood that the principles of the present invention can
be applied to diagnose disease in a body part without comparison to a
homolog. For example, the diagnostic WEVaI can be compared to a
population average, to the baseline value, or to some other standard to
diagnose disease.
Figure 5 shows a flowchart that illustrates the main steps 50 utilized by
system 10 to diagnose the possibility of disease in a body part. The first
part
of the procedure is preparatory and establishes standard or idealized
CA 02461972 2004-03-23
-23-
baselines for a typical body part and results are stored in the database to be
used as a reference for numerous subjects. At step (51 ), the baseline body
part is represented with a grid of finite elements. The grid can be two-
dimensional, or three-dimensional. Next, at step (52), nil current injections
are simulated to yield a database (54) of impedances and associated voltage
gradients. These steps may be repeated to collect several typical sets of data
depending on the size, body fat, or some other characteristic of the subject
or
the body part. This concludes the preparatory part. The subject-specific part
of the procedure is described next. At step (56) a plurality of electrodes is
applied to the body part, such as a breast and, at step (57), the plurality of
electrodes measure impedance of the body part between electrode pairs. At
step (58), a diagnostic is computed at each of the finite elements, the
diagnostic being a function of the measured impedance and the values of
impedance and gradients from the database. Subsequently, at step (60), the
diagnostic is utilized to diagnose the possibility of disease in the body
part.
Referring to Figures 6A and 6B, sample results in the form of two gray
scale plots are shown illustrating the value of the system and method of the
present invention in diagnosing breast cancer. In Figures 6A and 6B, the right
breast 72 and the left breast 74 are represented in the frontal plane as two
circular plots, with darkness of gray increasing as the homologous difference
of the diagnostic becomes more profound. This patient had an invasive ductal
adenocarcinoma in the mid outer right breast. To generate these circular
plots, each breast was represented by a circle with a 2D grid of finite
CA 02461972 2004-03-23
-24-
elements. In Figures 6A and fiB, the finite elements comprising the grid are
not shown.
The quantity I~ frt~ ~ - ~feft?I , as calculated by the comparator 26 for
homologous elements is, by convention, plotted on the side having the larger
WEVaI; i.e., on the right breast for elements where ~f~~,t ~ > ~feft J (Figure
6A)
and on the left breast where ~feR~ >tf,;~ ~ (Figure 6B). These differences are
scaled in the figure to the maximum level of black. Sixteen different levels
of
gray are presented, and some contrasting has been added to emphasize
areas where the differences are highest. However, none of these scaling
methods appreciably influenced the results. As can be seen in Figure 6B, the
shading in the normal left breast 74 is uniform (the light-most shade),
indicating that for this subject ~f,;~,t ~ > (f eft J everywhere.
Different computer systems can be used to implement the method for
diagnosing disease in a body part. The computer system can include a
monitor for displaying diagnostic information using one of several visual
methods. In one embodiment, the method can be implemented on a 2 GHz
PentiumT"" 4 system with 512 MB RAM.
It should be understood that various modifications and adaptations
could be made to the embodiments described and illustrated herein, without
departing from the present invention, the scope of which is defined in the
CA 02461972 2004-03-23
-25-
appended claims. For example, although emphasis has been placed on
describing a system for diagnosing breast cancer, the principles of the
present
invention can also be advantageously applied to other diseases of other body
parts. These body parts need not have a homolog. Also, although the main
measured electrical property described herein is impedance, it should be
understood that other electrical properties, such as functions of the
electrical
impedance, may also be used in accordance with the principles of the present
invention.
