U.S. patent application number 14/072480 was filed with the patent office on 2014-02-27 for method and apparatus using magnetic resonance imaging for cancer identification.
This patent application is currently assigned to Oregon Health & Science University. The applicant listed for this patent is Oregon Health & Science University. Invention is credited to Wei Huang, Xin Li, William D. Rooney, Charles S. Springer, Ian J. Tagge, Jingang Xu.
Application Number | 20140058249 14/072480 |
Document ID | / |
Family ID | 42129181 |
Filed Date | 2014-02-27 |
United States Patent
Application |
20140058249 |
Kind Code |
A1 |
Li; Xin ; et al. |
February 27, 2014 |
METHOD AND APPARATUS USING MAGNETIC RESONANCE IMAGING FOR CANCER
IDENTIFICATION
Abstract
Embodiments provide a Magnetic Resonance Imaging (MRI) technique
and optionally software--collectively referred to as the
"shutter-speed" model--to analyze image data of cancer patients.
Embodiments provide a minimally invasive, yet precisely accurate,
approach to determining whether tumors are malignant or benign by
distinguishing the characteristics of contrast reagent activity in
benign and malignant tumors. Exemplary embodiments provide MRI
measured biomarkers for tumor malignancy determination, effectively
eliminating or limiting the false positives suffered by existing
MRI techniques.
Inventors: |
Li; Xin; (Beaverton, OR)
; Springer; Charles S.; (Portland, OR) ; Rooney;
William D.; (Lake Oswego, OR) ; Huang; Wei;
(Lake Oswego, OR) ; Xu; Jingang; (Hillsboro,
OR) ; Tagge; Ian J.; (Portland, OR) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Oregon Health & Science University |
Portland |
OR |
US |
|
|
Assignee: |
Oregon Health & Science
University
Portland
OR
|
Family ID: |
42129181 |
Appl. No.: |
14/072480 |
Filed: |
November 5, 2013 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13125485 |
Apr 21, 2011 |
8605980 |
|
|
PCT/US2009/043201 |
May 7, 2009 |
|
|
|
14072480 |
|
|
|
|
61110404 |
Oct 31, 2008 |
|
|
|
61171411 |
Apr 21, 2009 |
|
|
|
Current U.S.
Class: |
600/420 ;
600/410 |
Current CPC
Class: |
A61B 5/4312 20130101;
G01R 33/56366 20130101; G01R 33/5601 20130101; G01R 33/4828
20130101; A61B 5/055 20130101; G01R 33/5608 20130101 |
Class at
Publication: |
600/420 ;
600/410 |
International
Class: |
A61B 5/055 20060101
A61B005/055; G01R 33/56 20060101 G01R033/56; A61B 5/00 20060101
A61B005/00; G01R 33/48 20060101 G01R033/48 |
Goverment Interests
GOVERNMENT INTERESTS
[0002] This invention was made with Government support under Grant
Nos. RO1-NS40801 and RO1-EB00422 awarded by The National Institutes
of Health. The Government has certain rights in the invention.
Claims
1. A method, comprising: obtaining a K.sup.trans value of a tissue
with an MRI device using a shutter-speed model (SSM); obtaining a
K.sup.trans value of the tissue with the MRI device using a
standard model (SM); determining a difference between the SSM
K.sup.trans value and the SM K.sup.trans value; and classifying the
tissue based on a determined difference between SSM K.sup.trans
value and the SM K.sup.trans value.
2. The method of claim 1, wherein the tissue is classified as
either a benign tumor or a malignant tumor.
3. The method of claim 1, wherein obtaining a K.sup.trans value
using SSM differs from obtaining a K.sup.trans value using SM in
that obtaining a K.sup.trans value using SSM includes a factor
reflecting equilibrium water exchange effects.
4. The method of claim 1, wherein classifying the tissue based on a
determined difference between SSM K.sup.trans value and the SM
K.sup.trans value comprises determining whether the difference is
above or below a defined threshold.
5. The method of claim 4, wherein a difference above the threshold
indicates the tissue is malignant, and wherein a difference below
the threshold indicates the tissue is benign.
6. The method of claim 4, wherein determining whether the
difference is above or below a defined threshold comprises
determining whether the difference is above or below a threshold at
or between 0.020 min.sup.-1 and 0.030 min.sup.-1.
7. The method of claim 4, wherein determining whether the
difference is above or below a defined threshold comprises
determining whether the difference is above or below 0.025
min.sup.-1.
8. The method of claim 1, wherein the K.sup.trans values of the
tissue are obtained utilizing dynamic contrast enhanced magnetic
resonance imaging (DCE-MRI).
9. The method of claim 8, further comprising obtaining a k.sub.ep
value for the tissue using DCE-MRI; establishing a two dimensional
plot of K.sup.trans and k.sub.ep for the tissue; establishing a
circle on the plot centered at an origin point of the plot, wherein
a radius of the circle defines a threshold; and classifying the
tissue based on whether the tissue plot point is greater than or
less than the threshold.
10. The method of claim 9, wherein a tissue plot point greater than
the threshold indicates the tissue is malignant.
11. The method of claim 9, wherein a tissue plot point less than
the threshold indicates the tissue is benign.
12. The method of claim 1, wherein obtaining a K.sup.trans value of
a tissue using SSM further comprises incorporating an interstitial
CR transverse relaxivity factor to address transverse relaxation
neglect present in SM.
13. A method for breast cancer screening, comprising: obtaining a
mammogram of breast tissue to determine whether the tissue has or
is suspected of having an occult lesion; obtaining a dynamic
contrast enhanced magnetic resonance image (DCE-MR image) of the
breast tissue determined to have or be suspected of having an
occult lesion; circumscribing a region of interest from the DCE-MR
image identifying breast tissue being or suspected of being an
occult lesion, wherein the region of interest is classified by
obtaining a K.sup.trans value of a tissue with an MRI device using
a shutter-speed model (SSM); obtaining a K.sup.trans value of the
tissue with an MRI device using a standard model (SM); determining
a difference between the SSM K.sup.trans value and the SM
K.sup.trans value; and classifying the tissue based on a determined
difference between SSM K.sup.trans value and the SM K.sup.trans
value.
14. A magnetic resonance imaging device, comprising: an image
capture component; and a processor configured to obtain a
K.sup.trans value of a tissue with the MRI device using a
shutter-speed model (SSM); obtain a K.sup.trans value of the tissue
with the MRI device using a standard model (SM); determine a
difference between the SSM K.sup.trans value and the SM K.sup.trans
value; and classify the tissue based on a determined difference
between SSM K.sup.trans value and the SM K.sup.trans value.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] The present application is a continuation of U.S. patent
application Ser. No. 13/125,485, filed Apr. 21, 2011, which is a
U.S. National Phase Application under 35 U.S.C. 371 of PCT
International Application No. PCT/US2009/043201, filed May 7, 2009,
which claims priority to U.S. Provisional Patent Application No.
61/171,411, filed Apr. 21, 2009, entitled "DCE-MRI Water Signal
Analysis for Improved Cancer Identification" and to U.S.
Provisional Patent Application No. 61/110,404, filed Oct. 31, 2008,
entitled "MRI Biomarker for Cancer Identification," the entire
disclosures of which are hereby incorporated by reference in their
entirety.
TECHNICAL FIELD
[0003] Embodiments herein relate to identification of cancer, and,
more specifically, to methods and apparatus using magnetic
resonance imaging for cancer identification.
BACKGROUND
[0004] Screening for breast cancer represents one of modern
medicine's success stories. However, the continued large fraction
of false positives in current diagnostic protocols often leads to
biopsy/pathology procedures that cause considerable pain, anxiety,
healthcare cost, and possibly increased malignancy risk, but which
are potentially avoidable. To address this problem, there have been
recent calls for the increased use of magnetic resonance imaging
(MRI) in breast screening.
[0005] The problems associated with false positive results are not
unique to breast cancer screening. Other cancers suffer from large
numbers of false positive results, causing significant stress as
well as often requiring additional costly and painful procedures to
confirm or deny the initial results.
BRIEF DESCRIPTION OF THE DRAWINGS
[0006] Embodiments will be readily understood by the following
detailed description in conjunction with the accompanying drawings.
Embodiments are illustrated by way of example and not by way of
limitation in the figures of the accompanying drawings.
[0007] FIG. 1 illustrates the pharmacokinetic modeling scheme for
DCE-MRI in accordance with various embodiments. The three general
compartments for contrast reagent (CR) and for water (blood,
interstitium, and parenchymal cytoplasmic) are illustrated--though
not in relative proportions to their volume fractions (v.sub.b,
v.sub.e, and v.sub.i). The pertinent chemical equilibria and their
unidirectional rate constants are indicated as well.
[0008] FIG. 2 illustrates K.sup.trans values obtained by a Standard
Model and by a Shutter-Speed Model in accordance with embodiments
herein.
[0009] FIG. 3 illustrates delta K.sup.trans results of a 77 lesion
data set (from 74 patients).
[0010] FIG. 4 illustrates a heat map analysis of a region of
interest (ROI), as delineated at the left.
[0011] FIGS. 5a, 5b, 5c, and 5d illustrate a sagittal,
fat-suppressed breast DCE-MRI image (FIG. 5a) containing a
malignant invasive ductal carcinoma (IDC) tumor (circled
contrast-enhanced region-of-interest). Pharmacokinetic K.sup.trans
parametric maps of the tumor, generated by the Standard Model
(FXL-constrained) and two members of the Shutter-Speed Model family
(FXR-allowed) and (SXR-allowed), are shown in FIG. 5b, FIG. 5c, and
FIG. 5d, respectively.
[0012] FIG. 6 illustrates 2D scatter plots of (a) the Standard
Model, and (b) the Shutter-Speed Model (FXR-a) results. The
ordinates measure the K.sup.trans and the abscissae the k.sub.ep
parameters. The black circles mark the positions for regions of
interest (ROIs) of lesions that were found by biopsy/pathology to
have large malignant fractions, while the triangles are those for
lesions found to be solely benign. An outlier is plotted in insets
c and d. Dashed concentric quarter-circles are drawn with radii of
0.19 and 0.23 min.sup.-1. The points for two patients are marked as
gray circles with black cores. These represent lesions with only
very small malignant fractions.
[0013] FIG. 7 illustrates a 1D scatter plot. The ordinate,
.DELTA.K.sup.trans, is [K.sup.trans (SSM)-K.sup.trans (SM)]: SSM is
FXR-a and SM is FXL-c. The values for the lesion ROIs of all 22
subjects are shown. Those proven malignant are given as filled
black circles (these include the two FIG. 6 gray circles with black
cores), while those found solely benign are indicated with
triangles. The group mean .DELTA.K.sup.trans values are indicated
with open and filled black squares on the right. Error bars
represent (SD) values within each category. One malignant lesion
outlier is plotted in an inset, and is excluded from the SD
calculation. The horizontal cut-off line drawn at 0.024 min.sup.-1
cleanly separates the two lesion groups.
[0014] FIG. 8 illustrates how the K.sup.trans (volume fraction CR
transfer rate constant product, top) and v.sub.e (extracellular,
extravascular space, EES, volume fraction, bottom) fitting results
would change if increasing interstitial .sup.1H.sub.2O T.sub.2*
quenching is assumed.
[0015] FIG. 9a (inset) shows a transverse pelvic DCE image slice
(anterior up/inferior perspective, .about.34 seconds post CR
injection) of a research subject. Two ROIs are indicated within the
prostate gland: one in an area of retrospectively-confirmed
prostate cancer, left; and the other in contralateral
normal-appearing prostate tissue, right. FIG. 9a plots the arterial
input function obtained from an ROI in a femoral artery. Its
magnitude was adjusted using a custom-written numerical approach
and an obturator muscle ROI for reference tissue. The time-course
from the first-pass was used to estimate blood volume fraction.
Color-matched tissue data time-courses (points) and representative
fittings (curves) are seen in FIG. 9b.
[0016] FIG. 10 illustrates an article of manufacture in accordance
with an embodiment herein.
DETAILED DESCRIPTION OF DISCLOSED EMBODIMENTS
[0017] In the following detailed description, reference is made to
the accompanying drawings which form a part hereof, and in which
are shown by way of illustration embodiments that may be practiced.
It is to be understood that other embodiments may be utilized and
structural or logical changes may be made without departing from
the scope. Therefore, the following detailed description is not to
be taken in a limiting sense, and the scope of embodiments is
defined by the appended claims and their equivalents.
[0018] Various operations may be described as multiple discrete
operations in turn, in a manner that may be helpful in
understanding embodiments; however, the order of description should
not be construed to imply that these operations are order
dependent.
[0019] The description may use perspective-based descriptions such
as up/down, back/front, and top/bottom. Such descriptions are
merely used to facilitate the discussion and are not intended to
restrict the application of disclosed embodiments.
[0020] For the purposes of the description, a phrase in the form
"A/B" or in the form "A and/or B" means (A), (B), or (A and B). For
the purposes of the description, a phrase in the form "at least one
of A, B, and C" means (A), (B), (C), (A and B), (A and C), (B and
C), or (A, B and C). For the purposes of the description, a phrase
in the form "(A)B" means (B) or (AB) that is, A is an optional
element.
[0021] The description may use the terms "embodiment" or
"embodiments," which may each refer to one or more of the same or
different embodiments. Furthermore, the terms "comprising,"
"including," "having," and the like, as used with respect to
embodiments, are synonymous.
[0022] In various embodiments, methods, apparatuses, and systems
using magnetic resonance imaging for cancer identification are
provided. In exemplary embodiments, a computing device may be
endowed with one or more components of the disclosed apparatuses
and/or systems and may be employed to perform one or more methods
as disclosed herein.
[0023] Embodiments herein provide a Magnetic Resonance Imaging
(MRI) technique and optionally newly developed
software--collectively referred to as the "shutter-speed" model--to
analyze image data of cancer patients. Embodiments provide a
minimally invasive, yet precisely accurate, approach to determining
whether tumors are malignant or benign. Exemplary embodiments
provide MRI measured biomarkers for tumor malignancy determination,
effectively solving the false positive riddle from which current
MRI techniques suffer.
[0024] Although some embodiments throughout are described with
reference to breast cancer or prostate cancer, the methods and
apparatuses described herein may be utilized for other cancers,
such as brain, esophageal, leg osteosarcoma, etc. as well as for
any Dynamic-Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI)
analysis where water exchange effects are relevant, including
tissue differential/disease state analysis of the brain
(Alzheimers, MS, etc.), muscles (such as heart), etc., and
quantitative vascular phenotype mapping.
[0025] "Quantitative MRI" produces parametric maps of MR,
patho-physiological, and/or pharmacokinetic biomarker properties.
The DCE-MRI sub-category is particularly significant because it
applies to a wide pathology range. In DCE-MRI, the T.sub.1-weighted
tissue .sup.1H.sub.2O MRI signal intensity is acquired before,
during, and after the (usually) bolus injection of a hydrophilic,
paramagnetic contrast reagent (CR). The CR passage through a tissue
region-of-interest (ROI) can cause a transient increase of the
longitudinal .sup.1H.sub.2O relaxation rate constant
[R.sub.1.ident.(T.sub.1).sup.-1] with consequent elevated MR
steady-state signal intensity. This elevation may be identified on
the MR image.
[0026] In DCE-MRI, the neglect of intercompartmental water exchange
kinetics considerations can lead to systematic errors in parameters
extracted by quantitative analyses. Examples here are the
compartmental water mole fractions defining tissue spaces.
Therefore, DCE-MRI is also a sub-category of in vivo MR "molecular
imaging"--mapping the distribution and/or activity of molecules in
living tissues.
[0027] In essence, in embodiments, the CR plays the role of the
nuclear medicine radioactive tracer. However, in nuclear medicine,
the tracer is detected directly (by its radioactivity in
disintegrations per second (dps)--the amount of tracer present in
the tissue, but compartmental localization is not intrinsic to the
signal). In contrast, the MRI CR is detected indirectly, via its
interaction with water and effect on the nature of tissue
.sup.1H.sub.2O relaxation (so the water interaction with the CR is
what is directly traced). Beneficially, the CR is not radioactive.
Also, MRI involves no ionizing radiation.
[0028] Affecting the recovery of longitudinal .sup.1H.sub.2O
magnetization (i.e., in the magnetic field direction) requires
(transient) water CR molecular interaction, as depicted in FIG. 1.
The three major loci for tissue water, the cytoplasmae, the
interstitium, and the blood, are indicated with subscripts i, o (or
e), and b (p, for plasma), respectively. There are water binding
equilibria depicted in each compartment in which the CR is thought
to enter. The compartmental volume fractions are designated as
v.sub.i, v.sub.e, and v.sub.b, respectively, though the relative
areas in FIG. 1 are not proportional.
[0029] The CR and water molecules are never equally distributed in
tissue. Therefore, the only way that most water (cytoplasmic) can
access CR is via exchange equilibria across cytolemmae and blood
vessel walls. These are indicated in FIG. 1 with the unidirectional
rate constants, k.sub.oi, k.sub.io, and k.sub.po, k.sub.op,
respectively. In existing methods, tracer pharmacokinetic models
are applied directly to MRI data--such methods are referred to here
as the Standard Model (SM). However, this results in the constraint
that all inter-compartmental equilibrium water exchange processes
be treated as if infinitely fast (k.sub.oi+k.sub.io.fwdarw..infin.,
and k.sub.po+k.sub.op.fwdarw..infin.). This is not valid, and the
assumption may effectively "short circuit" MRI determination of CR
compartmentalization--the pharmacokinetic essence. In accordance
with embodiments herein, the incorporation of equilibrium water
exchange MR effects into pharmacokinetic derivation is referred to
herein as the Shutter-Speed Model (SSM). This is accomplished by
allowing k.sub.oi+k.sub.io and k.sub.po+k.sub.op to be finite.
[0030] The SM assumes that water exchange between cells and/or
blood and the interstitial spaces is effectively infinitely fast
(in the fast exchange limit--FXL). However, when CR is passing
through the tumor tissue, the water exchange systems can depart
from this fast exchange limit due to the interaction with the CR
(and therefore enter into a fast-exchange regime--FXR). This
happens for both benign and malignant tumors; however, the exchange
difference between FXL and FXR, as far as K.sup.trans is concerned,
is significantly greater for malignant tumors as opposed to benign
tumors. For benign tumors, the exchange difference is typically
below 0.025 min.sup.-1, whereas for malignant tumors, the exchange
difference is typically above 0.025 min.sup.-1. This
differentiating line provides a threshold against which the
obtained values may be compared to classify a tumor or tissue
sample. In embodiments, a threshold may be established at an
exchange difference (delta K.sup.trans) of 0.02 to 0.03
min.sup.-1.
[0031] While a single threshold is mentioned above, in embodiments,
more than one delta K.sup.trans threshold may be established. For
example, a first threshold may be established that is intended to
include all or a substantial percentage of the malignant tissues
above the threshold. A secondary threshold may be established at a
lower delta K.sup.trans value with an intention of included all
true positive indicators. However, a lower threshold may introduce
a larger number of false positives. For a tissue having a delta
K.sup.trans value residing between the first and second threshold,
additional analysis may be utilized to further classify the
tissue.
[0032] In an embodiment, the shutter-speed model (SSM) accounts for
the FXR (therefore including equilibrium exchange effects when the
CR passes through) and thus is better able to pick up the "leaky
blood vessel" effect which is common in malignant tumors. At a
maximum level of CR in the interstitial space, an interstitial
water molecule in a benign lesion may typically encounter a CR
molecule an average of 60 times before it enters a cell, whereas in
a malignant tumor this may happen 260 times on average (4+ times as
often). If that difference is neglected (which the standard model
does), then it is sufficient to cause significant K.sup.trans (the
volume-weighted CR extravasation rate constant) underestimations in
malignant tumors. Because K.sup.trans values are greater for
malignant tissues than for benign tissues, if K.sup.trans is
underestimated, then it may make a malignant tumor seem benign
(false negative) or vice versa a benign tissue appear to be
malignant (false positive). The SSM model accounts for this
difference, and by using the delta K.sup.trans (change in
K.sup.trans) as well as the K.sup.trans to k.sub.ep comparison,
classification of tumors may be accomplished. In accordance with an
embodiment, k.sub.ep is the unidirectional CR intravasation rate
constant--it is K.sup.trans divided by the v.sub.e (the
extracellular, extravascular volume fraction available to the
contrast agent molecule). The pharmacokinetic analysis of DCE-MRI
data yields K.sup.trans and k.sub.ep.
[0033] In accordance with an embodiment, the difference in
K.sup.trans returned by SSM, as compared to the Standard Model
analyses, offers a very high degree of tumor differentiation (i.e.,
specificity). It is a measure of the shutter-speed effect, which is
disproportionally present and important in malignant tumors, that
permits differentiation of benign and malignant tumors. The
increased permeability of malignant tumor blood vessels exceeds a
threshold above which exchange kinetics become influential. This
amplification is measured by the delta K.sup.trans biomarker, and
accounts for the high SSM specificity.
[0034] In analyses of DCE-MRI data from patients with suspicious
breast lesions initially ruled positive by institutional screening
protocols, the SM K.sup.trans values for benign and malignant
lesions exhibit considerable overlap. The Shutter-Speed Model (SSM)
may allow for finite exchange kinetics thus agreeing with the SM
K.sup.trans value for each of the benign lesions. However, it
reveals that the SM underestimates K.sup.trans for each of the
malignant tumors in this population. FIG. 2 illustrates K.sup.trans
values obtained by both SM and SSM, and shows how SSM recognizes a
difference in K.sup.trans between benign and malignant tissue. The
fact that this phenomenon is unique to malignant tumors allows
their discrimination from the benign lesions, as validated by
comparison with gold standard pathology analyses of subsequent
biopsy tissue samples to which the MRI analyses were blinded.
Likewise, the SM overestimates k.sub.ep, particularly for the
benign tumors. Thus, incorporation of the SSM into the screening
protocols may preclude the need for the biopsy/pathology procedures
that otherwise would yield benign findings.
[0035] Thus, in embodiments, two binary classifiers have been
developed:
[0036] 1. "deltaK.sup.trans"--the change in K.sup.trans; thresholds
may be established with the goal/intention of including all true
positives. Thresholds may be established as desired to
distinguish/classify the tissues/tumors. In an embodiment, further
analysis may be conducted via a secondary mapping algorithm (plot
of (K.sup.trans vs. k.sub.ep) to allow for a second determination
with respect to those points that are somewhat unclear or fall
below a determined threshold.
[0037] 2. The use of 2D plots (K.sup.trans vs. k.sub.ep), where the
radius of a circle centered at the origin of the plot may be used
as a "binary classifier." In embodiments, the radius of the circle
may be used as a threshold to distinguish benign from malignant
tumors. Such a threshold may be established at approximately 0.2
min.sup.-1, for example, from 0.19 min.sup.-1 to 0.25
min.sup.-1.
[0038] In embodiments, an MRI examination aided by SSM analysis may
provide a clearer diagnosis and may be an intermediate step between
a mammographic scan and a biopsy intervention if breast cancer is
suspected from both the mammogram and the MRI results. Adding this
intermediate diagnostic step may greatly reduce or eliminate the
number of unnecessary (and possibly all) biopsy surgeries and also
reduce the pain, stress and expense for most patients.
[0039] It is important to note that the SSM is a generalization of
the SM. That is, the SM is but a special case of the SSM. Thus, if
the shutter-speed effect is negligible in any tissue, the program
will automatically perform a SM analysis. One can test this by
computationally constraining the SSM analysis to a SM form. If
there is no difference from the result obtained when the SSM
analysis is given free rein, then there is no significant
shutter-speed effect for that tissue. In the case of breast tumors,
this is the case for the K.sup.trans biomarker in only benign
lesions. However, there is a shutter-speed effect for the v.sub.e
biomarker in benign breast lesions, and it is about the same size
as in malignant tumors.
[0040] To test SSM, SSM was employed to analyze MR images of 22
women volunteers who had previously screened positive for breast
cancer by mammography and/or clinical examination. The
shutter-speed software operates by using a complex mathematical
formula to track the passage of injected contrast dye through a
tumor area. Contrast dyes are commonly used in medical imaging to
increase the visibility of tissue abnormalities.
[0041] When viewed through the shutter-speed analysis, the MRI data
suggested that only seven of the 22 women actually had malignant
tumors. These projections were later shown to be 100 percent
accurate after each of the study participants underwent subsequent
biopsies for pathology determinations. Typically, 75 percent of
mammographically-indicated biopsies yield negative pathology
results, meaning that an intermediate step such as an MRI
determination could greatly reduce or eliminate the number of
unnecessary biopsy surgeries.
[0042] This population study has been expanded to include 77 breast
tumors (in 74 patients) and, with the mapping provision for one
rare type of malignant tumor, maintains 100% specificity. FIG. 3
illustrates delta K.sup.trans results of the 74 patient data set
illustrating 77 lesions.
[0043] In addition, FIG. 3 illustrates the multiple threshold
concept described previously. The first threshold is intended to
capture all or a substantial percentage of the malignant tissues
above the threshold. The first delta K.sup.trans threshold may be
set, for example, between 0.02 min.sup.-1 and 0.03 min.sup.-1. The
secondary threshold is established at a lower delta K.sup.trans
value with an intention of included all true positive indicators.
The second delta K.sup.trans threshold may be set, for example,
between 0.01 min.sup.-1 and 0.02 min.sup.-1. For a tissue having a
delta K.sup.trans value residing between the first and second
threshold, additional analysis may be utilized to further classify
the tissue. Such analysis may include heat map analysis of regions
of interest to better classify the tissue.
[0044] FIG. 4 illustrates a heat map analysis of a ROI, as
delineated at the left. The three maps to the right show the
results of various analyses of the ROI. At the top, the SM (FXL-c)
image is shown which does not provide an indicator of malignancy.
The middle image represents the SSM (FXR-a) image, which indicates
some areas of interest. However, the lower image, representing the
delta K.sup.trans value, clearly outlines the particular areas of
concern within the ROI. Even though this lesion is fairly early
stage, the delta K.sup.trans analysis provides an indicator of the
tumor malignancy. While in some situations the identification of
the tumor malignancy may not result in treatment, the early
identification enables the tumor growth to be tracked over a period
of time.
[0045] For the more limited data set (22 patients), data were
obtained with consent from patients with positive mammographic
and/or clinical MRI reports from standard, institutional breast
cancer workups and protocols. All had MRI contrast-enhanced lesions
radiologically classified as BIRADS (Breast Imaging Reporting and
Data System) four (B-4, suspicious) or five (B-5, highly suggestive
of malignancy). Emphasizing practicability and robustness, the data
are of a rather routine clinical nature (and they were obtained at
two different institutions, with two different instruments, CRs,
etc.): the two different data acquisitions were not optimized for
DCE-MRI. For example, though the spatial resolution is reasonable,
the temporal resolution is not optimal. Of particular interest is
the fact that the adipose tissue --.sup.1H.sub.2C-- MR signal was
suppressed in the acquisitions at one institution, while at the
other institution it was not.
[0046] FIG. 5a shows the DCE pharmacokinetic image of sagittal
slice 16 (numbering from lateral to medial) of the left breast of a
52 year-old patient, obtained 2.6 minutes after CR injection. It
was acquired with adipose --.sup.1H.sub.2C-- suppression (required
in the institutional protocol). In contrast to those with no fat
suppression, this darker image shows glandular regions brighter
than fatty tissue. The ROI circumscribes the enhanced lesion
evident in this slice, subsequently found to be a malignant
invasive ductal carcinoma (IDC) by pathology analysis. Each of the
22 patients participated in a DCE-MRI acquisition subsequent to her
clinical mammography and/or MRI screening but prior to the biopsy
procedure and the pathology analysis.
[0047] Additional DCE-MRI acquisition details may be found in Li,
et al., Dynamic NMR Effects in Breast Cancer
Dynamic-Contrast-Enhanced MRI, PNAS, Vol. 105, No. 46, 17937-17942
(2008) (and Supporting Online Material), the entire disclosures of
which are incorporated by reference in their entirety. For each of
the 22 subjects, ROI DCE-MRI time-course data were analyzed from
one sagittal image slice (out of 16 to 40 per breast) that
exhibited a lesion to be subsequently biopsied. An ROI boundary was
manually drawn around the entire lesion in a pharmacokinetic image
showing near maximal enhancement (as in FIG. 5a). The patients are
enumerated in Table 1, below. The FIG. 5 images are from patient 3.
The DCE-MRI time-courses were each analyzed with several
pharmacokinetic models.
TABLE-US-00001 TABLE 1 K.sup.trans (min-.sup.1)
k.sub.ep(min.sup.-1) Patient BI- SM SSM SM SSM Number RADS (FXL-c)
(FXR-a) (FXL-c) (FXR-a) Pathology Report 1 B-4 0.073 0.147 0.389
0.249 DCIS, intermediate nuclear grade 2 B-4 0.110 0.180 0.452
0.447 IDC, histologic grade II/III; DCIS, intermediate nuclear
grade; DCIS .ltoreq.25% of total tumor mass 3 B-4 0.087 0.131 0.161
0.147 IDC present at the edge of the core 4 B-5 0.164 0.254 0.532
0.432 IDC, histologic grade II/III; DCIS, (.+-.0.028) (.+-.0.029)
intermediate nuclear grade; DCIS >25% of total tumor mass 5 B-5
0.559 1.63 1.795 2.966 IDC, histologic grade II/III (.+-.0.040)
(.+-.0.06) 6 B-5 0.145 0.185 0.506 0.308 IDC (.+-.0.020) 7 B-4
0.051 0.081 0.269 0.202 IDC, nuclear grade II, LCIS: moderately
differentiated IDC embedded within a larger benign LCIS 8 B-5 0.033
0.034 0.106 0.053 LCIS, SF (.+-.0.005) (.+-.0.006) 9 B-4 0.022
0.023 0.147 0.047 FC 10 B-4 0.051 0.058 0.306 0.155 FC 11 B-4 0.040
0.055 0.280 0.120 FC 12 B-4 0.062 0.077 0.397 0.188 Sclerosed
papillary lesion, LCIS 13 B-4 0.027 0.028 0.048 0.039 FC 14 B-4
0.030 0.034 0.229 0.096 ADH, SF 15 B-4 0.091 0.099 0.189 0.131
LCIS, SF, FC 16 B-4 0.078 0.087 0.188 0.130 LCIS. ADH 17 B-4 0.108
0.125 0.289 0.166 duct ectasia, ADH 18 B-4 0.060 0.066 0.133 0.090
SF, sclerosing adenosis 19 B-4 0.048 0.050 0.185 0.086 FA 20 B-4
0.026 0.028 0.174 0.066 FA (.+-.0.010) (.+-.0.005) 21 B-5 0.020
0.022 0.307 0.136 FA 22 B-4 0.016 0.016 0.436 0.078 FA IDC:
invasive ductal carcinoma; DCIS: ductal carcinoma in situ; LCIS:
lobular carcinoma in situ; SF: stromal fibrosis; FC: fibrocystic
changes; ADH: atypical ductal hyperplasia; FA: fibroadenoma.
[0048] For the patients/results presented in Table 1, ROI
boundaries around each lesion were separately drawn by each of two
independent investigators who were blinded to the pathology
results. The analyses of these ROI data were also conducted
independently by two investigators. The algebraic means of the
model parameters returned from each investigator's fitting were
computed lesion-by-lesion.
[0049] Each of the fittings neglects the small blood water proton
signal (.sup.1H.sub.2O.sub.b)--thus, these are "first generation"
versions. For this situation, the MR exchange system of interest is
that for equilibrium transcytolemmal water interchange (k.sub.oi
and k.sub.io, FIG. 1). The system's condition is given by the
comparison of the equilibrium kinetics, k=k.sub.oi+k.sub.io, with
the pertinent MR shutter-speed,
.tau..sup.-1.ident.|R.sub.1o-R.sub.1i|, where R.sub.10 and R.sub.1i
are the relaxation rate constants for the .sup.1H.sub.2O.sub.o and
.sup.1H.sub.2O.sub.i signals in the absence of exchange. Before CR
arrival, R.sub.1o.apprxeq.R.sub.1i and .tau..sup.-1<<k.
Though k is finite, and invariant throughout the DCE-MRI study, the
system is in the fast-exchange-limit (FXL): the kinetics appear
infinitely fast, and the measured tissue .sup.1H.sub.2O R.sub.1 is
single-valued. As stated above, the Standard Model assumes that the
system remains in the FXL throughout the CR bolus passage, so it is
referred to also as the FXL-constrained (FXL-c) model (see FIG.
5b). However, as the CR.sub.o concentration increases, R.sub.1o
becomes increasingly larger than R.sub.1i and .tau..sup.-1 at least
approaches the constant k value. For some period, the measured
R.sub.1 remains effectively single-valued, and this has been
defined to be the fast-exchange-regime (FXR). Admitting departure
from the FXL for the FXR may be referred to as FXR-allowed (FXR-a)
(see FIG. 5c). Further CR.sub.o increase may lead to the condition
where R.sub.1 is effectively double-valued: this is referred to as
the slow-exchange-regime (SXR). Admitting this is referred to as
SXR-allowed (SXR-a) (see FIG. 5d). For the cases here, the results
of FXL-c and FXR-a analyses are presented in Table 1. Careful
analyses with the SXR-a model suggest that it is incompatible with
these data--an example will be seen below. There are a number of
potentially variable parameters. For the SM (FXL-c) analyses, the
variables were K.sup.trans and v.sub.e, while for the SSM (FXR-a)
analyses, .tau..sub.i was also varied. In terms of the FIG. 1
notation, K.sup.trans=v.sub.ek.sub.ep=v.sub.bk.sub.pe, and
.tau..sub.i=k.sub.io.sup.-1. The values returned for K.sup.trans, a
measure of the rate of passive CR transfer across the vessel wall,
and k.sub.ep, the unidirectional rate constant for CR intravasation
(FIG. 1) are given in Table 1. Sample standard deviation measures
of parameter uncertainty from individual fittings are given for
some entries. These were determined by multiple Monte Carlo fitting
calculations. The K.sup.trans and k.sub.ep values for the malignant
tumors (top seven entries) are larger than those for the benign
lesions.
[0050] Table 1 indicates that the SM does not completely separate
the malignant tumors (top seven entries) from the benign lesions
with either the K.sup.trans or k.sub.ep parameters. However, the
SSM significantly increases K.sup.trans for every one of the
malignant lesions, and for none of the benign tumors, as compared
to the SM. Furthermore, though the SSM reduces k.sub.ep for both
malignant and benign lesions, it does this more for the benign
tumors. In embodiments, these changes allow discrimination between
the SSM and SM results.
[0051] Though neither of the parameters allows the construction of
perfect ROC (Receiver Operator Characteristic) plots, the SSM
K.sup.trans and k.sub.ep quantities come very close. These aspects
may be seen in the 2D parametric scatter plots of the K.sup.trans
(ordinate) and k.sub.ep (abscissa) values presented in FIG. 6. The
ROI values for lesions found by pathology analyses (Table 1) to be
solely benign are indicated with triangles, while those with major
malignant regions are shown as black circles. The two gray circles
with black cores also represent malignant tumors and are discussed
below. The results from the SM (FXL-c) analyses are seen in panel
a, while those from the SSM (FXR-a) determinations are shown in
panel b. The values for patient 5 are so large that they are shown
in inset panel c and inset panel d.
[0052] In comparing FIG. 6b with 6a, one can note especially the
upward movement (increasing K.sup.trans) of the circles and the
leftward movement (decreasing k.sub.ep) of the triangles, in going
from the SM to the SSM. This allows the almost complete separation
of these points in FIG. 6b, which is not achieved in any single
dimension of either panel. It is important to note that two of the
triangles represent B-5 lesions (Table 1): i.e., they were "highly
suggestive" false positives. Retaining 100% sensitivity (not
missing any malignant tumor), the PPV values for the SM
K.sup.trans, SM k.sub.ep, SSM K.sup.trans, and SSM k.sub.ep
dimensions are: 54%, 39%, 70%, and 70%, respectively. In the FIG.
6b SSM 2D plot, one can draw a dashed quarter-circle of radius 0.19
min.sup.-1, that also allows a 78% PPV.
[0053] Furthermore, consider the annular region between this and
the other concentric quarter-circle, of radius 0.23 min.sup.-1. The
only two malignant tumors (circles with dark cores within) are
those of patients 3 (upper) and 7 (lower). These are cases where
the malignant areas are quite small compared with the total tumor
area visualized in the biopsy specimen (Table 1). This means that
the analyses of whole-tumor ROI-averaged data cause a partial
volume dilution of the DCE-MRI parametric values. This can be seen
clearly in panels b and c of FIG. 5, which present K.sup.trans
parametric heat maps of the lesion of patient 3. In the SM (FXL-c)
and SSM (FXR-a) maps (panels b and c, respectively), a clear "hot
spot" is seen on the posterior lesion edge. The hot spot has
K.sup.trans values above 0.16 min.sup.-1 in the FXR-a map,
considerably elevated above the ROI-averaged magnitude (Table
1).
[0054] The hot regions of all seven malignant tumors in this
population have SSM K.sup.trans values exceeding 0.1 min.sup.-1.
Except for that of patient 17 (upper triangle in FIG. 6b annulus,
and which uniquely exhibits ductal dilation (Table 1)), this
exceeds the ROI-averaged SSM K.sup.trans values of any of the
fifteen benign lesions. With FIGS. 5b, 5c, and 5d, parametric maps
(heat maps) of four of the seven malignant tumors have been
presented in several publications referenced below. Some hot spots
can be as small as 2 mm in diameter. In another indication of
potential staging power, a plot (not shown) of "hotness" vs. area
of the SSM K.sup.trans hot spots in the malignant tumors of
patients 5, 6, and 7 demonstrates that these two independently
measured quantities are very highly positively correlated. The fact
that the SXR-a K.sup.trans map of the patient 3 lesion (FIG. 5d)
does not show increased values relative to the FXL-c map (FIG.
5b)--and in fact obliterates the hot spot--is an example of the
SXR-a model incompatibility with these data.
[0055] The K.sup.trans and k.sub.ep values are rather well
correlated in FIG. 6, particularly in panel b. The positions of the
panel a and b insets are placed with constant coordinate aspect
ratios. Thus, one can visually include the inset points in the
correlations. The slope of a line drawn through the points
represents the mean v.sub.e value of these lesions. Such a line for
FIG. 6b has a slope near 0.5.
[0056] These results suggest a potential breast cancer screening
protocol in accordance with an embodiment herein. The first step of
such a protocol would be a clinical examination and/or mammography.
A positive result (B-4 or B-5), or suspicion of a mammographically
occult lesion, would occasion referral for diagnostic MRI that
includes DCE. The radiologist can circumscribe an ROI from the DCE
image showing the greatest enhancement. Alternatively, this can be
automated (ex., Jim 4.0 software; Xinapse Systems; Thorpe
Waterville, UK). The computer can very quickly (few seconds)
conduct SM and SSM analyses on the mean ROI signal time-course data
and produce SSM K.sup.trans and k.sub.ep values, which can be
compared with 2D scatter plots such as those in FIG. 6b. If a
patient's point turns out to be in the annulus between the
quarter-circles in FIG. 6b, the radiologist could proceed to read
K.sup.trans parametric lesion maps made from the same DCR-MRI data,
though these require more computational time. Hot spots above 0.1
min.sup.-1 would be very suspicious for malignancy.
[0057] Some oncologists advocate a separate regimen for a malignant
ductal carcinoma in situ (DCIS) tumor, possibly simply following it
instead of immediate surgery, while others urge excision. The only
solely DCIS case in the discussed patient population is that of
patient 1. Her position in FIG. 6b is the black point closest to
the outer quarter-circle. In fact another concentric quarter-circle
of radius 0.3 min.sup.-1 would isolate this point. Its position
could be "followed" or tracked over a period of time to see if it
moves up and to the right. Inside the inner quarter-circle, most of
the benign LCIS lesions are found in the upper right sector, while
all of the FA lesions are found near the bottom.
[0058] In the analyses so far, pseudo-absolute parameter values
have been employed. The SSM success suggests that neglect of
equilibrium transcytolemmal water exchange effects may constitute
the most significant systematic error in Standard Model DCE-MRI
pharmacokinetic analyses.
[0059] For screening purposes, the most striking aspect of the
Table 1 and FIG. 6 results is that every one of the malignant tumor
ROI K.sup.trans values (dark circles) is clearly decreased by the
SM analysis, while every one of the benign lesion ROI values
(triangles) is not. This is seen even more clearly in FIG. 7, which
presents the 1D scatter plot for .DELTA.K.sup.trans
[.ident.K.sup.trans(SSM)-K.sup.trans(SM)]. There is a wide gap
between all seven of the dark circles [group mean, 0.06 min-1
(excluding the inset point)], and all 15 of the triangles. The
latter set clusters very near zero [group mean, 0.006 min-1]. A
clean cut-off line is drawn at 0.024 min.sup.-1. Since the only
difference between these two models is the allowance for the effect
on the NMR signal of finite equilibrium transcytolemmal water
exchange kinetics, the NMR shutter-speed effect, this suggests that
it is significant (for the K.sup.trans magnitude) with the
capillary wall permeability obtained for the vascular beds of only
malignant breast tumors. Thus, this is very encouraging that
analyses of DCE-MRI ROI data first with one pharmacokinetic model
and then with the other (which is still accomplished in only
seconds) can lead to extremely high specificity in cancer
screening. Here, the positive criterion of
.DELTA.K.sup.trans>0.025 min.sup.-1 yields 100% PPV.
[0060] Apparently, in the vascular beds of malignant breast tumors
only, the interstitial ("outside") CR concentration, (CR.sub.o),
transiently rises to sufficient values during the bolus passage and
the equilibrium transcytolemmal water exchange system transiently
departs the FXL to sufficient extent and/or for sufficient duration
to substantially invalidate the SM K.sup.trans determination. The
SSM interpretation is that, during the bolus passage through
malignant lesions, the relaxographic .tau..sup.-1 value for the
transcytolemmal water exchange process, |R.sub.1o-R.sub.1i|,
transiently approaches or exceeds that for the unchanging exchange
rate constant, k.sub.io+k.sub.oi, (in vivo studies are isothermal)
sufficiently for the system to enter at least the fast-exchange
regime (FXR), but probably not also the slow-exchange-regime (SXR).
R.sub.1o increases with CR.sub.o, while R.sub.1i remains constant.
This is a manifestation of the varying equilibrium competition for
interstitial water molecules between diamagnetic cytoplasmic spaces
and paramagnetic interstitial CR molecules (FIG. 1). Informative
estimates can be made by comparison of the Table 1 patients 8/4
benign/malignant lesion pair, with SSM K.sup.trans 0.034 and 0.254
min.sup.-1, respectively. For one of the SSM (FXR-a) fittings of
each, the (v.sub.e,.tau..sub.i) parameters returned are similar:
(0.60, 0.40 s), and (0.69, 0.39 s) for benign and malignant,
respectively. Thus, the unidirectional rate constants for water
cellular entry
[k.sub.oi.apprxeq.(v.sub.e.sup.-1-1).tau..sub.i.sup.-1] are similar
(1.7 and 1.2 s.sup.-1, respectively), constant, and not infinitely
large. However, before the arrival of interstitial CR.sub.o, the
transcytolemmal water exchange appears infinitely fast in the NMR
signal because .tau..sup.-1 is almost negligible. The interstitial
water molecules encounter no paramagnetic CR.sub.o molecules before
entering a diamagnetic cytoplasm. However, as CR.sub.o increases,
the rate constant for interstitial water CR encounter,
[(CR.sub.o)/(H.sub.2O.sub.o)].tau..sub.M.sup.-1, also increases
[.tau..sub.M.sup.-1=k.sub.M in FIG. 1]. While, for the benign
lesion CR.sub.o maximizes at 0.52 mM (at .about.7.5 minutes), this
is 1.6 mM (at .about.3.5 minutes) for the malignant tumor. Thus,
[(CR.sub.o)max/(H.sub.2O.sub.o)].tau..sub.M.sup.-1 values are 104
and 313 s.sup.-1 for the benign and malignant lesions,
respectively. The interstitial water concentration (H.sub.2O.sub.o)
was 50 M and the mean water lifetime on the CR, .tau..sub.M, was
10.sup.-7 s. At maximum CR.sub.o, an interstitial water molecule in
the benign lesion encounters a paramagnetic CR molecule on average
60 times (104/1.7) before it enters a diamagnetic cell; sufficient,
apparently, for the SM 40% v.sub.e underestimation. While in the
malignant tumor, this happens 260 times (313/1.2) on average; more
than four times as often. This is sufficient to cause significant
K.sup.trans underestimations if it is neglected.
[0061] For the expanded data set, a total of 74 patients underwent
clinical breast MRI protocols and had 77 contrast-enhanced lesions
(3 patients presented 2 lesions each) radiologically classified in
the BIRADS (Breast Imaging Reporting and Data System) 4 (B-4,
suspicious, n=67) or 5 (B-5, highly suggestive of malignancy, n=10)
categories based on lesion morphology and qualitative enhancement
kinetics assessment. These clinical interpretations led to biopsy
referrals. The research DCE-MRI data acquisitions were
IRB-approved. The data from 6 patients were collected as part of a
combined MRI/MRS protocol prior to excisional or core biopsy. Those
from the other 68 patients (71 lesions) were acquired during
clinically scheduled MRI-guided preoperative needle localization or
core biopsy procedures, just before needle insertions.
[0062] The study was conducted at 1.5 T using a body transmit and a
four- or seven-channel phased-array bilateral breast receive RF
coils. A 3D SPGR pulse sequence was used to acquire 12-20 serial
sagittal image volume sets continually, spatially covering the
whole breast with the suspicious lesion to be biopsied. Other
parameters included 10.degree. or 30.degree. (for the 6 patients)
flip angle, 2-5 ms TE, 6-9 ms TR, 3 mm section thickness, 20-24 cm
FOV. Depending on the breast size, 16-36 image sections were
acquired for each set, resulting in inter-sampling intervals of
13-42 seconds. At the start of the second volume set acquisition,
Gd CR was delivered intravenously [0.1 mmol/kg at 2 mL/s]. ROIs
circumscribing the enhanced lesion and within an axillary artery
produced the tumor signal intensity and arterial input function
(AIF) time-courses, respectively. Three reliable individual AIFs
were measured, which were interpolated with an empirical expression
(3) and averaged to generate a mean AIF. The tumor ROI and mean AIF
signal time-courses were then subjected to both SM and
(fast-exchange-regime-allowed) SSM analyses, which were blinded
from the pathology. Receiver-operating-characteristic (ROC) curves
were used to evaluate pharmacokinetic parameter diagnostic
accuracies, and the areas under the curve (AUCs) were compared
using a Bootstrap nonparametric test.
[0063] Upon pathology, only 18 lesions (10 B-4 and 8 B-5) were
found malignant and the other 59 (57 B-4 and 2 B-5) benign. Though
the clinical MRI protocol sensitivity is 100% (no false negatives),
its PPV is only 23%. The SSM K.sup.trans ROC AUC (0.973) is
significantly (p=0.032) greater than that for the SM K.sup.trans
(0.929). Similar results were obtained for other strong biomarkers:
k.sub.ep (=K.sup.trans/v.sub.e, the unidirectional CR intravasation
rate constant) [SSM AUC=0.960, SM AUC=0.861, p=0.006] and
[(K.sup.trans).sup.2+k.sub.ep.sup.2].sup.1/2 [SSM AUC=0.970, SM
AUC=0.887, p=0.009]. Maintaining 100% sensitivity, the diagnostic
specificities of the SM ROI K.sup.trans, k.sub.ep, and
[(K.sup.trans).sup.2+k.sub.ep.sup.2].sup.1/2 are 47%, 42%, and 51%,
while those for the corresponding SSM parameters are 76%, 61%, and
75%, respectively; each biomarker used as a binary classifier. The
SM and SSM v.sub.e ROC curve AUCs are 0.555 and 0.615,
respectively, suggesting that v.sub.e is a poor diagnostic
marker.
[0064] FIG. 3 (discussed partially above) plots ROI delta
K.sup.trans for all lesions. Note the ordinate scale break. Each
column represents one pathology category (from left to right): 1)
invasive ductal carcinoma (IDC)/ductal carcinoma in situ (DCIS)
mixture, 2) IDC/invasive lobular carcinoma (ILC) mixture, 3) IDC,
4) DCIS, 6) IDC/lobular carcinoma in situ (LCIS) mixture, and 9)
ILC, for the malignant group (circles); 5) tubular adenoma, 7)
LCIS, 8) atypical lobular hyperplasia, 10) atypical ductal
hyperplasia, 11) stromal fibrosis, 12) benign parenchyma, 13)
fibrocystic changes, 14) papillary lesions, 15) miscellaneous
benign conditions, 16) fibroadenomatoid changes, and 17)
fibroadenoma, for the benign group (triangles). The categories are
ranked roughly in order of decreasing mean delta K.sup.trans from
left to right. Consistent with the previous smaller population
study, the delta K.sup.trans biomarker represents the strongest
binary classifier for benign and malignant group separation, with
its ROC AUC=0.990, and 88% specificity for 100% sensitivity.
[0065] The SSM DCE-MRI ROI pharmacokinetic parameters consistently
perform better than those from SM DCE-MRI and the commonly used
clinical MRI protocols for benign and malignant discrimination
within this group of 77 suspicious breast lesions. If the simple
ROI delta K.sup.trans analyses had been integrated into clinical
practice, as many as 52 benign lesions (68% of the total
population) could have been spared the biopsy procedures. As
expected from the earlier study, the malignant lesions cluster
almost exclusively on the left of FIG. 3, while the benign lesions
are almost all to the right--the axes are independent. The solid
cut-off line value, delta K.sup.trans=0.028 min.sup.-1, is very
close to that for 100% specificity in the smaller population. It
yields only one false positive (the sole tubular adenoma) and one
false negative (the sole ILC) lesion. A more lenient, dashed
cut-off line can be drawn at delta K.sup.trans=0.012 min.sup.-1 to
avoid any false negative and still incur only 14 benign biopsies.
But, even these might be avoided. The likely reason for a malignant
lesion ROI delta K.sup.trans to fall between the solid and dashed
cut-off lines is because of partial volume averaging in the ROI
analyses. Consistent with this, the ILC had the very large value of
5 cm as the greatest enhanced ROI dimension. Its pixel-by-pixel SSM
K.sup.trans map (not shown) features hot spots (K.sup.trans>0.18
min.sup.-1) only in the posterior rim region. Though these are
diluted by a very large area of small K.sup.trans values in the
ROI, they confirm the lesion as malignant. This suggests that delta
K.sup.trans or SSM K.sup.trans maps (parametric heat maps) should
be made when an ROI delta K.sup.trans falls between the solid and
dashed lines.
[0066] Other analysis methods may be used to further distinguish
the data, such as points falling between delta K.sup.trans
thresholds. For example, the median K.sup.trans difference, delta
(median K.sup.trans) [SSM median K.sup.trans-SM median
K.sup.trans], may be plotted (ordinate) vs. the change in maximum
histographic probability (amplitude) (abscissa), delta Amp [SSM
amplitude-SM amplitude]. Such an operation indicates a significant
negative linear correlation (Pearson correlation=-0.82, p=0.0018)
for the benign lesions, while the malignant lesions exhibit an
almost orthogonal correlation. The ILC (identified for example in
FIG. 3) cannot be distinguished from the benign group with simple
ROI delta K.sup.trans analyses. However, using quadratic
discrimination analysis, the benign and malignant lesions can be
completely separated (100% sensitivity and 100% specificity) by the
solid partition curve with no misclassification.
[0067] Though the ROI delta K.sup.trans biomarker achieves high
specificity for benign/malignant breast lesion discrimination, the
partial volume averaging effects of ROI analyses can cause overlap
in ROI pharmacokinetic parameter values, and thus prevent clearer
separation of the two groups. Pharmacokinetic parametric mapping
and histogram analyses thus may further improve discrimination.
Such analyses are especially important when the lesion ROI
biomarker value falls in the vicinity of a binary classifier
cut-off value. Thus, it is beneficial to acquire DCE-MRI data with
sufficient SNR, since this ensures reliable pixel signal
time-course curve fitting. The negative linear correlation of the
benign lesions and the orthogonal behavior of the malignant lesions
are quite interesting. Compared to malignant lesions that can have
noticeable median K.sup.trans increases (shutter-speed (SS)
histographic shifts) without significant histographic maximum
probability changes (SS broadening), the areas in benign lesions
where increased blood vessel CR permeability incurs SS effects, if
any, are smaller. Considerable SS histographic broadening is
associated with even minuscule SS histographic shifting.
[0068] Further details regarding the materials and methods used
with respect to various embodiments described herein as well as
details regarding some of the MRI data acquisitions and analyses
may be found in Li, et al., Dynamic NMR Effects in Breast Cancer
Dynamic-Contrast-Enhanced MRI, PNAS, Vol. 105, No. 46, 17937-17942
(2008) (and Supporting Online Material); Huang, et al., The MR
Shutter-Speed Discriminates Vascular Properties of Malignant and
Benign Breast Tumors In Vivo, PNAS, Vol. 105, No. 46, 17943-17948
(2008); Li, et al., Shutter-Speed Analysis of Contrast Reagent
Bolus-Tracking Data: Preliminary Observations in Benign and
Malignant Breast Disease, Magn. Reson. Med., 53:724-729 (2005); and
Yankeelov, et al., Evidence for Shutter-Speed Variation in CR
Bolus-Tracking Studies of Human Pathology, NMR Biomed., 18:173-185
(2005), the entire disclosures of which are hereby incorporated by
reference.
[0069] In accordance with embodiments herein, certain steps may be
taken, even in the clinical setting, to improve the precision, the
accuracy, and/or the diagnostic richness of the SSM DCE-MRI
pharmacokinetic parameters. Such modifications may, for example,
decrease the random error scatter in the FIGS. 6 and 7 point
clusters. This may allow further discrimination of pathology
sub-types.
[0070] The DCE-MRI time-course acquisitions discussed herein were
prescribed for radiological considerations and were truncated.
Increasing this period would likely improve accuracy and precision
of the benign lesion parameters. For these ROIs, the maximum
R.sub.1 value is rarely reached in the no more than seven minutes
usually allowed. This is the likely source of abnormally large
v.sub.e values for some benign tumors. Increasing the period to 15
minutes may help define the shape of the time-course, even for
malignant tumors.
[0071] The DCE-MRI acquisitions for the data described herein were
not particularly exchange sensitive. Even so, exchange effects seem
to facilitate very high discrimination of malignant from benign
breast tumors.
[0072] The tissue R.sub.10 values (the pre-CR .sup.1H.sub.2O
longitudinal relaxation rate constants) may be mapped, and not
simply assumed as they were herein. Individual AIFs may be used as
well. A reference tissue method, or an automated AIF determination
(ex., Jim 4.0 software; Xinapse Systems; Thorpe Waterville, UK) may
be used.
[0073] Increased temporal resolution may be achieved without
sacrificing spatial resolution or signal-to-noise. Parallel RF
excitation/acquisition may be useful for achieving such increased
temporal resolution. With good definition of the DCE time-course
first-pass leading edge, the second generation SSM (BALDERO (Blood
Agent Level Dependent and Extravasation Relaxation Overview))
analysis, which accounts for blood .sup.1H.sub.2O signal
pharmacokinetic behavior, may be used to also determine v.sub.b and
k.sub.bo values. It is anticipated that tumor v.sub.b values will
have significant diagnostic value. Furthermore, v.sub.bk.sub.bo is
the transendothelial water permeability coefficient surface area
product, P.sub.WS', where S' is the total capillary bed surface
area. The ratio P.sub.WS'/P.sub.CRS' would be the intensive
property P.sub.W/P.sub.CR. The value of the CR permeability
coefficient surface area product (P.sub.CRS') may be factored from
the K.sup.trans parameter using the blood flow value, which may
also be determined from DCE-MRI data.
[0074] The DCE-MRI pharmacokinetic images may also be spatially
registered to correct for patient motion.
[0075] Image acquisition without --.sup.1H.sub.2C-- suppression may
yield signal intensities much more amenable to precision parametric
mapping. The maps require sufficient acquisition contrast-to-noise
ratio because pixel-by-pixel analytical modeling is more
susceptible to noise. However, care must be taken to avoid
contamination of .sup.1H.sub.2O by unsuppressed
--.sup.1H.sub.2C--.
[0076] In embodiments, the shutter-speed model may be enhanced by
adding a factor for putative T.sub.2* (transverse relaxation)
signal quenching. In an embodiment, there is provided a direct
application of a T.sub.2* reduction factor to the interstitial
water signal in the Ernstian MR steady-state DCE-MRI model
expression. Assuming the greatest T.sub.2* reduction will return
K.sup.trans and v.sub.e values for the tumor region of interest
about 35% and 15% greater, respectively, than one would find when
ignoring this effect. For normal-appearing tissues, these are 11%
and 17% greater, respectively. Thus, applying the factor further
distinguishes normal tissue from the tumor ROI. FIG. 8 illustrates
this relationship.
[0077] The SXR-a SSM includes T.sub.2* neglect and therefore
underestimates K.sup.trans and v.sub.e to the extent that there is
a disproportionate relaxation of compartmental water signals.
Embodiments herein provide a way of testing to see if the blood and
interstitial water signals have been edited from the detected
signal (that is, SXR-a is inappropriate).
[0078] DCE-MRI pharmacokinetic modeling usually ignores potential
.sup.1H.sub.2O signal reduction due to transverse relaxation
(T.sub.2*) effects. Most clinical DCE-MRI applications employ a
contrast reagent (CR) dose of 0.1 mmol/kg which may produce a blood
plasma CR concentration above 5.0 mM at its peak during the bolus
passage. Here, using exemplary prostate DCE-MRI data, a potential
T.sub.2* effect on DCE-MRI model parameter values is described, by
using a water exchange ("shutter-speed") model along with a
simplified factor to account for putative T.sub.2* signal
quenching.
[0079] Prostate .sup.1H.sub.2O MRI data were acquired with a
Siemens TIM Trio (3T) system under an IRB approved protocol. RF
transmitting was through the whole body coil and RF receiving was
with a combination of Spine Matrix and flexible Body Matrix RF
coils. The DCE-MRI sequence employed a 3D TurboFLASH sequence with
a 256*144*16 matrix size and a 360*203 mm.sup.2 field of view,
resulting in an in-plane resolution of 1.4*1.4 mm.sup.2. Other
parameters are: slice thickness: 3 mm; TR/TE/FA: 5.42 ms/1.56
ms/15.degree., imaging intersampling interval: 4.16 seconds. Any
T.sub.2*-induced signal reduction is assumed to be proportional to
[exp(-(r.sub.2*(CR)+R.sub.20)TE)], applying to the .sup.1H.sub.2O
signal from the CR-occupied compartment. For the data here, the
most influential CR-containing compartment is the prostate
interstitium. Thus, r.sub.2* and CR represent the interstitial CR
transverse relaxivity and concentration, respectively. Since
susceptibility effects cross compartmental boundaries, surely
r.sub.2* also has a contribution from capillary blood plasma CR.
This T.sub.2*-reduction factor is then directly applied to the
interstitial .sup.1H.sub.2O signal in the Ernstian MR steady-state
DCE-MRI model expression. Parameter uncertainties were determined
with sets of Monte Carlo simulations carried out for each
ROI-averaged .sup.1H.sub.2O signal with increasing T.sub.2*
quenching accounted for by choosing an increasing r.sub.2* value
(mM.sup.-1s.sup.-1): 0 (no quenching), 5 (a literature value), 20
(an estimated blood plasma value at 3 T), or 40. For each r.sub.2*
and each ROI data set, 200 simulation runs were performed with
Gaussian noise (p=0, a=0.08) directly added to the normalized ROI
data time-course. This resulted in a simulated time-course with a
signal-to-noise ratio (SNR) slightly better than that from a single
pixel. Random initial guess values were evenly distributed within
the parameter space for each simulation fitting.
[0080] FIG. 9a inset shows a transverse pelvic DCE image slice
(anterior up/inferior perspective, .about.34 seconds post CR
injection) of a research subject. Two ROIs are indicated within the
prostate gland: one in an area of retrospectively-confirmed
prostate cancer, left; and the other in contralateral
normal-appearing prostate tissue, right. FIG. 9a plots the arterial
input function obtained from an ROI in a femoral artery. Its
magnitude was adjusted using a custom-written numerical approach
and an obturator muscle ROI for reference tissue. The time-course
from the first-pass (includes the initial peak) was used to
estimate blood volume fraction. Color-matched tissue data
time-courses (points) and representative fittings (curves) are seen
in FIG. 9b.
[0081] FIG. 8 shows how the K.sup.trans (volume fraction CR
transfer rate constant product, top) and v.sub.e (extracellular,
extravascular space, EES, volume fraction, bottom) fitting results
would change if increasing interstitial .sup.1H.sub.2O T.sub.2*
quenching is assumed. With K.sup.trans values this large, the
algorithm is effectively a two-site (interstitium/cytoplasmae)
exchange model, and the T.sub.2*-induced signal reduction is
applied to only the EES signal. As noted above, assuming the
greatest T.sub.2* reduction (r.sub.2*=40 mM.sup.-1s.sup.-1) will
return K.sup.trans and v.sub.e values for the tumor ROI about 35%
and 15% greater, respectively, than one would find ignoring this
effect. For the normal-appearing tissue, these are 11% and 17%
greater, respectively. Conversely, the usual literature analysis
includes transverse relaxation neglect (by effectively assuming
r.sub.2*=0) and thus underestimates K.sup.trans and v.sub.e to the
extent that there is disproportionate relaxation of compartmental
.sup.1H.sub.2O signals.
[0082] The analysis used here is based on an inherently three-site
model, but multi-step recursive fittings would eventually return a
zero (within error) blood volume fraction (v.sub.b) for the tumor
tissue. This is not because v.sub.b is actually zero, but only
because it is indeterminate due to the very CR-permeable capillary
wall. The blood .sup.1H.sub.2O signal makes a contribution
indistinguishable from that of the EES. Thus, it may be better to
use an only two-site model. For consistency, the same two-site
model is also used for the normal appearing tissue ROI. The current
analysis is conservative in estimating EES signal
T.sub.2*-quenching effects. Interestingly, however, the extracted
parameters move exactly in the direction seen comparing analyses
with the fast-exchange-regime (FXR)-allowed two-site shutter-speed
model with the slow-exchange-regime (SXR)-allowed version. The
former neglects a distinguishable interstitial .sup.1H.sub.2O
signal contribution, which is reduced by exchange and may also be
at least partially T.sub.2*-quenched. For a tumor blood volume
estimation using DCE-MRI with extravasating CR, it is prudent to
use a lower CR dose.
[0083] Any one or more of various embodiments previously discussed
may be incorporated, in part or in whole, into a computing device
or a system. A suitable computing device may include one or more
processors for obtaining/receiving data, processing data, etc. One
or more of the processors may be adapted to perform methods in
accordance with various methods as disclosed herein. A computing
device may also include one or more computer readable storage
media.
[0084] Any one or more of various embodiments as previously
discussed may be incorporated, in part or in whole, into an article
of manufacture. In various embodiments and as shown in FIG. 10, an
article of manufacture 1000 may comprise a computer readable medium
1010 (a hard disk, floppy disk, compact disk, etc.) and a plurality
of programming instructions 1020 stored in computer readable medium
1010. In various ones of these embodiments, programming
instructions 1020 may be adapted to program an apparatus, such as
an MRI device or a processor within or separate from an MRI device,
to enable the apparatus to perform one or more of the
previously-discussed methods.
[0085] In an embodiment, a computing device/system may be
configured to receive MR images through any of a variety of
communication schemes (wired or wireless), to analyze the data as
described herein to classify the tissue that was the subject of the
MR image, and to display/transmit the results. The computing
device/system may be configured to receive MRI data from an
integrated MRI device or from a separate MRI device in
communication electronically. The computing device/system may then
display the analysis results on an integrated display, or may send
the results to a separate computing device, using any suitable
electronic communication mechanism, for separate display and
potentially further analysis.
[0086] Although certain embodiments have been illustrated and
described herein, it will be appreciated by those of ordinary skill
in the art that a wide variety of alternate and/or equivalent
embodiments or implementations calculated to achieve the same
purposes may be substituted for the embodiments shown and described
without departing from the scope. Those with skill in the art will
readily appreciate that embodiments may be implemented in a very
wide variety of ways. This application is intended to cover any
adaptations or variations of the embodiments discussed herein.
Therefore, it is manifestly intended that embodiments be limited
only by the claims and the equivalents thereof.
* * * * *