U.S. patent application number 10/598764 was filed with the patent office on 2011-05-26 for automated neuroaxis (brain and spine) imaging with iterative scan prescriptions, analysis, reconstructions, labeling, surface localization and guided intervention.
Invention is credited to Judd M. Storrs, Kenneth L. Weiss.
Application Number | 20110123078 10/598764 |
Document ID | / |
Family ID | 34975781 |
Filed Date | 2011-05-26 |
United States Patent
Application |
20110123078 |
Kind Code |
A9 |
Weiss; Kenneth L. ; et
al. |
May 26, 2011 |
Automated Neuroaxis (Brain and Spine) Imaging with Iterative Scan
Prescriptions, Analysis, Reconstructions, Labeling, Surface
Localization and Guided Intervention
Abstract
Automated spine localizing, numbering and autoprescription
system enhances correct location of diseased or injured tissue,
even allow multi-spectral diagnosis. Externally located this tissue
is facilitated by an integrated self adhesive spatial reference and
skin marking system that is designed for a variety of modalities to
include MRI, CT, SPECT, PET, planar nuclear imaging, radiography,
XRT, thermography, optical imaging and 3D space tracking. The
device ranges from a point localizer to a more multifunctional and
complex grid/phantom system. The specially designed spatial
reference(s) is affixed to an adhesive strip with corresponding
markings so that after applying the unit to the skin/surface and
imaging, the reference can be removed leaving the skin
appropriately marked. The localizer itself can also directly adhere
to the skin after being detached from the underlying strip. A spine
autoprescription process performs image analysis that is able to
identify vertebrae and discs even in the presence of
abnormalities.
Inventors: |
Weiss; Kenneth L.;
(Cincinnati, OH) ; Storrs; Judd M.; (Cincinnati,
OH) |
Prior
Publication: |
|
Document Identifier |
Publication Date |
|
US 20070223799 A1 |
September 27, 2007 |
|
|
Family ID: |
34975781 |
Appl. No.: |
10/598764 |
Filed: |
September 11, 2006 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
PCT/US05/08311 |
Mar 11, 2005 |
|
|
|
10598764 |
Sep 11, 2006 |
|
|
|
60552332 |
Mar 11, 2004 |
|
|
|
Current U.S.
Class: |
382/131 |
Current CPC
Class: |
Y10T 428/24744 20150115;
G06T 2207/10088 20130101; G06T 7/11 20170101; G06T 2207/20221
20130101; Y10T 428/24 20150115; G06T 2207/20101 20130101; A61B
5/0033 20130101; G06T 2207/30016 20130101; A61B 5/0263 20130101;
A61B 5/4836 20130101; G06T 7/0012 20130101; B60R 25/00 20130101;
H04L 2012/40273 20130101; A61B 5/0042 20130101; G06K 2209/055
20130101; A61B 5/0036 20180801; G06T 2207/10016 20130101; G06T
2207/10072 20130101; G06T 2207/30012 20130101 |
Class at
Publication: |
382/131 |
International
Class: |
G06K 9/00 20060101
G06K009/00 |
Claims
1. An apparatus for identifying and labeling spinal structures in a
medical diagnostic image of a patient, comprising: a memory
configured to receive the medical diagnostic images; a program
stored in the memory and operatively configured to detect a
plurality of voxels in the image as candidate spinal structures, to
apply at least one spinal structure constraint to identify a subset
of the plurality of voxels as a series of spinal structures, and to
label the series of spinal structures as a selected specified one
of a cervical, thoracic, lumbar vertebral structures; and a
processor in communication with the memory to perform the
program.
2. The apparatus of claim 1, wherein the program is further
operatively configured to identify a selected seed vertebral
structure and to identify adjacent vertebral structures by locating
a longest chain of voxels and analyzing spacing and quantity of
voxels in the longest chain.
3. The apparatus of claim 1, wherein the program is further
configured to define putative spine structures by defining a search
region, applying intensity thresholds, applying additional disc
constraints, and identifying the longest chains.
4. The apparatus of claim 3, wherein the program is further
configured to combine data and then search along a reconstructed
path neighborhood for local maxima.
5. The apparatus of claim 4, wherein the program is further
configured to make a determination whether twenty three spinal
discs have been selected, and if not to search for an additional
discs based on estimated interdisc distance for each level.
6. The apparatus of claim 4, wherein the program is further
configured to search for an additional disc by extending a line
from the longest chain and analyze for an additional disc.
7. The apparatus of claim 4, wherein the program is further
configured to search for additional discs by extending the search
line, drawing additional parallel lines to the extended line and
analyze for additional discs.
8. The apparatus of claim 4, wherein the program is further
configured to search for an additional disc by analyzing an
elongate space between adjacent discs and analyze for an additional
disc by adjusting a search constraint.
9. The apparatus of claim 4, wherein the program is further
configured to apply optimized Gaussian filters to the search
part.
10. The apparatus of claim 4, wherein the program is further
configured to search for auto-prescribed additional imaging planes
and sequences.
11. The apparatus of claim 1, wherein the program is further
configured to analyze a selected spinal structure to diagnose
osteoporosis.
12. A method for performing a medical diagnostic imaging scan of a
patient, comprising: placing a longitudinally unique opaque spinal
coil on external to a spine of a patient; performing a scout scan;
identifying and labeling on diagnostic scans each vertebral body of
the spin; autoprescribing a portion proximate to a vertebral body
for a detailed scan; identifying a unique longitudinal position of
the spinal coil proximate to a surgical site contained within the
autoprescribed portion; and inserting a therapeutic instrument
localized by the spinal coil to the surgical site.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] The present application claims the benefit of the U.S.
Provisional Patent Application Ser. No. 60/552,332, filed 11 Mar.
2004, the disclosure of which is hereby incorporated by reference
in its entirety.
FIELD OF THE INVENTION
[0002] The present invention relates, in general, to medical
diagnostic imaging devices that perform scout scans for
localization and autoprescription.
BACKGROUND OF THE INVENTION
[0003] Diagnostic imaging of the spine of a patient is often useful
for identifying disease or an injury to the spine itself or as a
readily locatable landmark for other tissues. Unfortunately, human
error may occur due to the variability in the patient population or
due to an oversight. The mistake may arise in incorrectly labeling
vertebrae and discs in a diagnostic image. The mistake may arise in
incorrectly visually identifying the corresponding vertebrae under
the skin before performing a surgical or therapeutic (e.g.,
radiation) treatment. The mistake may arise in improperly
identifying a normal, benign, or malignant condition because an
opportunity is missed to correctly correlate information from a
plurality of imaging systems (e.g., a type of tissue may be
determined if an MRI and a CT image could be properly correlated
and analyzed).
[0004] With regard to spine prescription, a number of complications
exist that necessitate having an extensively trained clinician
identify the imaged vertebrae. For instance, the quality of the
diagnostic image may vary depending on the source and type of
imaging modality. The presented image volume provided may not
include the top and bottom vertebrae. Vertebrae and discs may not
be adequately captured in the image due to congenital defect,
disease, injury or surgery. The individual in question may have an
atypical number of mobile pre-sacral vertebrae, either more or less
than 23. Further, the spacing and curvature of the individual's
spine may be rather exceptional.
[0005] Even if the vertebrae and discs may be accurately identified
in the diagnostic image, it is often helpful to be able to obtain a
one-to-one correspondence between the readily visible and markable
skin/surface and underlying structures or pathology detectable by a
variety of imaging modalities. This may facilitate clinical
correlation, XRT, image guided biopsy or surgical exploration,
multimodality or interstudy image fusion, motion
correction/compensation, and 3D space tracking.
[0006] Current methods, (e.g. bath oil/vitamin E capsules for MRI),
have several limitations including single image modality utility
requiring completely different and sometimes incompatible devices
for each modality, complicating the procedure and adding potential
error in subsequent multimodality integration/fusion. They require
a separate step to mark the skin/surface where the localizer is
placed and when as commonly affixed to the skin by overlying tape,
may artifactually indent/compress the soft tissue beneath the
marker or allow the localizer to move, further adding to potential
error. Sterile technique is often difficult to achieve.
Furthermore, it may be impossible to discriminate point localizers
from each other or directly attain surface coordinates and
measurements with cross sectional imaging techniques. In regards to
the latter, indirect instrument values are subject to significant
error due to potential InterScan patient motion, nonorthogonal
surface contours, and technique related aberrations which may not
be appreciated as current multipurpose spatial reference phantoms
are not designed for simultaneous patient imaging.
[0007] The trend is to take and digitally store lots of data on a
patient, including MR and CT images. You want to both compare each
patient's data to his/her own data, and "pools" of data from other
people. Little problem: How do you make sense of pictures taken at
different times, using different types of machines and different
actual machines, for the same or different people? That's what Dr.
Weiss accomplishes with his techniques for the skull:
well-characterized fixed reference points. He proposes something
similar for the spine. Nothing "automatic" exists today and there
are no real standards for how to characterize points of reference
on the skull, let alone the spine.
[0008] Limited coverage, resolution and contrast of conventional
MRI localizers coupled with a high prevalence of spinal variance
make definitive numbering difficult and may contribute to the risk
of spinal intervention at the wrong level. Only 20% of the
population exhibit the classic 7 cervical, 12 thoracic, 5 lumbar, 5
sacral, and 4 coccygeal grouping. For instance, 2-11% of
individuals have a cephlad or caudal shift of lumbar-sacral
transition, respectively resulting in 23 or 25 rather than the
typical 24 mobile presacral vertebrae. Numbering difficulties are
often heightened in patients referred for spine MRI. Such patients
are more likely than the general population to have anomalies,
acquired pathology, or instrumentation that distorts the appearance
of vertebrae and discs. Moreover, these patients are often unable
to lie still within the magnet for more than a short period of time
due to a high prevalence of back pain and spasms. Resultant
intrascan motion confounds image interpretation and interscan
motion renders scan coordinates and positional references
unreliable.
[0009] While data remains somewhat limited, various authors report
an approximately 2-5% incidence of wrong level approach spinal
intervention, with most cases involving the lower lumbar
interspaces. Such surgical misadventures may lead to needless pain
and suffering, as well as contribute to accelerating medical
malpractice costs. The first multi-million dollar jury verdict for
such a wrong level approach was awarded in 2002.
[0010] Although several research techniques have been described to
automate spine image analysis, to the authors' best knowledge, none
has successfully addressed the need for accurate and unambiguous
numbering. Computer characterization of a vertebrae or disc is of
limited clinical value if that structure can not be accurately
identified and named.
[0011] Consequently, a significant need exists for an improved
approach to localizing and autoprescribing through multi-modal
quick scans of the brain and/or spine. Furthermore, there is a need
for enhancing personal medicine with a method of aligning skull and
spine images.
[0012] Once one image set is autoprescribed, it would be further
beneficial to correlate with other types of imaging modalities that
are also autoprescribed. One advantage is that calculations of
changes over time for the same patient may quickly identify injury
or disease. Another advantage is that different spectral emissions
illicit different information about a tissue. Correlating between a
plurality of imaging modalities, if a common tissue structure can
be localized for each, may enable autodiagnosis as to whether the
tissue is normal, benign or malignant. Consequently, it would be of
a further advantage to extend spine autoprescription across
multiple sources of diagnostic images.
BRIEF SUMMARY OF THE INVENTION
[0013] The present invention addresses these and other problems in
the prior art by providing an apparatus and method of localizing a
spinal column of a patient with robust automated labeling of
vertebrae across a population and across different imaging
modalities facilitating autoprescription and follow-on therapeutic
procedures. Thereby, human error in incorrectly identifying a
vertebrae in an image, and thus mislocating a surgical site, is
avoided.
[0014] In one aspect of the invention, an apparatus processed a
medical diagnostic image of a patient's torso by identifying voxels
of appropriate size to be putative spinal structures. Then disc
constraints are applied to identify a long chain of spinal
structures that are then labeled.
[0015] In another aspect of the invention, this processing is in
conjunction with a localized coil placed on the torso of the
patient that provides an external reference correlated to the
identification and labeling, enabling accurate insertion or aiming
of therapeutic treatments.
[0016] By virtue of the foregoing, an entire spine can be
effectively surveyed with sub-millimeter in-plane resolution MRI in
less than one minute. All cervical-thoracic-lumbar vertebrae and
discs can be readily identified and definitively numbered by visual
inspection. All cervical-thoracic-lumbar vertebrae and discs can be
readily identified and definitively numbered by semi-automated
computer algorithm. Rapid technique should facilitate accurate
vertebral numbering, improve patient care, and reduce the risk of
surgical misadventure. Coupled with an integrated head and spine
array coil, rapid computer automated iterative prescription and
analysis of the entire neuro-axis may be possible.
[0017] These and other objects and advantages of the present
invention shall be made apparent from the accompanying drawings and
the description thereof.
BRIEF DESCRIPTION OF THE FIGURES
[0018] The accompanying drawings, which are incorporated in and
constitute a part of this specification, illustrate embodiments of
the invention, and, together with the general description of the
invention given above, and the detailed description of the
embodiments given below, serve to explain the principles of the
present invention.
[0019] FIG. 1 is a diagram of an automated spinal diagnostic
system.
[0020] FIG. 2 is a flow diagram of a spine identification sequence
or operations or procedure for the automated spinal diagnostic
system of FIG. 1.
[0021] FIG. 3 is a diagram of a projection function of adjacent
objects (.sub.p and (.sub.q represent the angles between the line
connecting candidate disc p and q through their centroid and the
major axis of disc p and q respectively, wherein
0<(.sub.p<90.degree. and (.sub.p<45.degree. and
(.sub.p<45.degree. let d.sub.c be the value of d for any
candidate disc in cervical-thoracic spine region and d.sub.L in the
thoracic-lumbar spine, then 6 mm<d.sub.c<80 mm and 8
mm<d.sub.L<100 mm.
[0022] FIG. 4 is a diagram of distance constraint chains with a
cluster, k, is part of a disc chain if its closest superior
neighbor has k as its closest inferior neighbor and k's closest
inferior neighbor has (k) as its closest superior neighbor.
[0023] FIG. 5 is a 7-slice sagittal MRI projected image volume
having a 35 cm FOV top half and
[0024] FIG. 6 is a 35 cm FOV bottom half illustrating typical
search regions wherein voxels exceeding intensity threshold are
depicted with those meeting additional disc constraints are
highlighted as putative disks and connected by a curved line
through their centroids.
[0025] FIG. 7 is a combined sagittal image depicting search paths
parallel to the curved line of FIG. 6 connecting a centroid of a
C2-3 disc with longest disc chains from top and bottom half images
(FIGS.). Dots correspond to local maxima along these paths.
[0026] FIG. 8 is a sagittal image through the entire spine with all
intervetebral discs auto-labeled with labeling of vertebrae omitted
for clarity with three-dimensional (3-D) coordinates generated by
an algorithm providing a means for discs or vertebral bodies to be
labeled in any subsequent imaging plane providing no gross
inter-scan patient movement.
[0027] FIG. 9 a sagittal image through the entire spine of a
patient with 23 mobile/presacral vertebrae with correct
auto-labeling of first 22 interspaces.
[0028] FIG. 10 a sagittal image through the entire spine of a
patient with 25 potentially mobile presacral vertebrae with correct
auto-labeling of the first 23 interspaces.
[0029] FIG. 11 a sagittal image through the entire spine of a
patient with surgically fused L4-5 interspace and associated
artifact from a metallic cage with correct labeling of all 23
interspaces including a good approximation of the L4-5 disc.
[0030] FIG. 12 a sagittal image through the entire spine of a
patient with vertebral planus of T-10 mislabeled due to a less
robust disc discrimination process.
[0031] FIG. 13 is a sagittal image through the entire spine of the
patient of FIG. 12 with correctly labeled vertebrae due to a more
robust disc discrimination process including Gaussian filters.
[0032] FIGS. 14A-14I are diagrams of a point localizer; FIG. 14A
depicts a frontal view of the point localizer affixed to fabric;
FIG. 14B depicts a reverse side of the point localizer of FIG. 14A;
FIG. 14C is a perspective view of the point localizer and
underlying fabric affixed to the skin; FIG. 14D is an enface view
of the fabric with corresponding marking affixed to the skin; FIG.
14E is an enface view of the localizer affixed to skin; FIG. 14F is
a diagram view of a port integrated into a tubular ring; FIG. 14G
is a frontal view of a modified ring shaped localizer affixed to
fabric with additional markings; FIG. 14H is a frontal view of the
fabric in FIG. 14G with the localizer removed; FIG. 14I is a
frontal view of a multilocalizer sheet demonstrating the adhesive
backing and overlying fabric with localizers removed.
[0033] FIGS. 15A-15F illustrate a cross-shaped grid configuration;
FIG. 15A is an enface view of the grid with modified square as the
central hub and uniquely positioned rows of tubing radiating along
the vertical and horizontal axes; FIG. 15B is a schematic of axial
cross sections acquired at representative distances from the
horizontal axis; FIG. 5C demonstrates the underlying marked fabric
with the superimposed tubing in FIG. 15A removed; FIG. 15D is a
variant of FIG. 15A with modified ring serving as a central hub;
FIG. 15E depicts a limb fixation ring and angulation adjuster; and
FIG. 15F depicts a radiopaque grid with underlying ruled fabric
strips removed.
[0034] FIG. 16A is an enface view of the grid/phantom configuration
with tubular lattice, overlying a diagonally oriented slice
indicator, and underlying a partially adhesive fabric with markings
and perforations; FIG. 16B is a schematic cross section of a
representative axial section of the grid/phantom configuration of
FIG. 16A.
[0035] FIGS. 17A-17B are diagrams of localizers in a packaged strip
or roll, regularly spaced at 5 cm or other intervals.
[0036] FIGS. 18A-18B are a lattice localizer having tube diameters
varied to identify unique locations.
[0037] FIGS. 19A-19D are depictions of a fall-spin grid localizer
and spinal coil.
DETAILED DESCRIPTION OF THE INVENTION
Spine Localization, Automated Labeling, and Data Fusion Diagnostic
System.
[0038] In FIG. 1, an automated spinal diagnostic system 10 includes
a diagnostic imaging system 12 (e.g., MRI, CT) that is used to
image a torso of a patient 14 that is advantageously covered by a
skin/surface marking system 16 that serves as an integrated
multimodality, multi-functional spatial reference. The diagnostic
imaging system 12 may include scanning of the skull 18, the full
spine 20, and pelvic bones 22. The diagnostic imaging system 12
serves as an automated MRI technique that rapidly surveys the
entire spine providing accurate definitive numbering of all discs
and vertebrae. In the particular illustrative version, the entire
spine can be effectively surveyed with sub-millimeter in-plane
resolution MRI in less than 1 minute. C-T-L vertebrae and discs can
be readily identified and definitively numbered by visual
inspection or semi-automated computer algorithm ("ASSIST").
[0039] Correctly identifying each vertebrae and disc in the spine
20 is complicated in certain instances when the skull 18 and/or the
pelvic bones 22 are not imaged. In some instances, a vertebrae or
disc (depending on whether CT or MRI is being used respectively)
will fail to image properly, depicted at 24. In addition, the
highly predominant number of twenty-three vertebrae may not be the
case for certain individuals, such as twenty-four as depicted.
Having the correct vertebrae references may be important in
localizing a suspicious lesion 26, such as for a later therapeutic
procedure, represented as a radiation device 28.
[0040] The diagnostic imaging system 12 may derive a sagittal image
28 of the torso 14 from a volume CT scan 30. Alternatively, the
diagnostic imaging system 12 may produce an upper cervical-thoracic
sagittal image 32 and a lower thoracic-lumbar sagittal image 34,
such as from MRI. A spine autoimage processor 40, which may be a
process hosted by the diagnostic imaging system 12 or by a remote
device, performs a number of subprocesses to correctly identify and
label the spine 20.
[0041] In an illustrative version of the diagnostic imaging system
12, MRI studies were performed on a clinical 1.5T magnet with
standard 4-channel quadrature array 6-element spine coil and
surface coil correction algorithm. Contiguous two-station 35 cm FOV
sagittal FGRE sequences were pre-programmed, providing full
cervical, thoracic and lumbar (C-T-L) spine coverage. Combined
sagittal FOV of 70 cm., 7 sections, L15-R15, 4 mm skip 1 mm;
512.times.352, 1 nex, TR 58, TE 2.0, 30.degree. flip, BW 15.6; 21
sec.times.2 stations=42 sec. To facilitate and standardize
auto-prescriptions, a line was drawn on the spine coil for the
technologists to landmark (set as scanner's 0 coordinate) rather
than have them use a superficial anatomic feature. The coil has a
contoured head/neck rest assuring grossly similar positioning of
the cranial-cervical junction of each patient relative to this
landmark. The semi-automated disc detection and numbering algorithm
of the spine image processor 40 was iteratively developed, tested
and refined on batches of consecutive de-identified patient studies
and results compared to neuroradiologist assignments. The spine
image processor 40 was implemented in MATLAB 7.
[0042] In block 41, a computer algorithm is hosted on the spine
image processor for the identification and labeling of disc and
vertebrae from auto-prescribed sagittal MRI or sagittal
auto-reformatted CT data, described in greater detail below. This
information may be advantageously provided to an automated spine
image analysis algorithm 43 that further characterizes each disc
and/or vertebrae level. Alternatively or in addition, this
information from block 41 may be advantageously provided to an
auto-prescription algorithm 45 for additional image sequences,
utilizing the identified spinal landmarks as references. Further,
the additional processes 43, 45 may exchange information with each
other, such as detained analysis and diagnosis of a particular
vertebrae in block 43 being enabled by auto-prescribed detailed
images in block 45.
[0043] These analyses performed by the spine image processor 40 in
the illustrative version key upon one or more sources of
information that identify the top vertebrae.
[0044] In particular, in block 46, a top vertebrae is identified,
which may be automated in block 48 by interfacing with an automated
cranium identification system, such as described in U.S. patent
application Ser. No. 10/803,700, "AUTOMATED BRAIN MRI AND CT
PRESCRIPTIONS IN TALAIRACH SPACE" to Dr. Weiss, filed 18 Mar. 2004,
the disclosure of which is hereby incorporated by reference in its
entirety. Alternatively, logic may be incorporated into the spine
image processor 40 wherein a spine algorithm in block 50 recognizes
a top vertebrae. As a further alternative and as fall back option
should automated means fail, in block 52 a technologist input is
received to identify the top vertebrae.
[0045] The neuroradiologist could readily visualize and
definitively number all C-T-L levels on ASSIST localizers. 6/50
patients had a congenital shift in lumbar-sacral transition; 3 with
23 mobile pre-sacral vertebrae and 3 with 25 mobile pre-sacral
vertebraeBased upon usual of manual placement of a single seed in
the C2-3 interspace for accurate identification and numbering of
the other 22 discs, in the illustrative version with 50/50 cases
performed by the spine autoprescription processor 40. The automated
disc detection and numbering algorithm was concordant with
neuroradiologist assignments in 50/50 (100%) of cases.
[0046] With a labeled disc image 64, correct relative location to
other imaged tissue of interest, depicted at 65 may be used for
further diagnostic procedures or therapeutic intervention. For
instance, in block 66 with an ability to correlate images taken
with the same type of imaging modality at different times, growth
progression of a suspicious lesion or changes due to an intervening
injury may be compared between images. In addition, images taken
with different imaging modalities may be cross referenced to
perform multi-spectral diagnoses (block 68), wherein information on
a type of tissue may be gained by its different responses to
various types of electromagnetic spectra. With tissue diagnosis
complete, in block 70 may correctly orient a therapeutic agent,
such as the radiation device 28 or a guided minimally invasive
surgical instrument (not shown). Alternatively, for an externally
oriented procedure, a surgeon may reference either the relative
location to a known spinal constituent (e.g., vertebrae) and/or
reference the skin/surfacing marking system 16.
[0047] In an illustrative implementation of the spine image
processor 40 of FIG. 1, Internal Review Board (IRB) approval for
the research study was obtained. As part of a revised thoracic
spine clinical MRI protocol instituted at an outpatient imaging
facility, patients received a rapid total spine ASSIST localizer
with pre-set parameters. All studies were performed on a 1.5T GE
Excite MRI system with standard 4-channel, 6-element quadrature
spine coil and surface coil correction algorithm. Contiguous
two-station 35 cm FOV sagittal FGRE sequences were pre-programmed,
providing full cervical, thoracic and lumbar (C-T-L) spine
coverage. Combined sagittal FOV was 70 cm., 7 sections obtained at
each station, L15-R15, 4 mm skip 1 mm; 512.times.352, 1 nex, TR 58,
TE 2.0, 30.degree. flip, BW 15.6; 21 sec.times.2 stations=42 sec.
These ASSIST studies were de-identified in consecutive batches and
copied to CD for subsequent off-line review, computer algorithm
development and testing. A semi-automated disc detection and
numbering algorithm was iteratively developed and results compared
to neuroradiologist assignments.
[0048] The first batch of 27 cases was initially run with an
algorithm 100 developed using previously obtained, non surface-coil
corrected ASSIST images, The first step, we input cervicothoracic
(top half), and thorariclumbar (bottom half) spines. A threshold
and a median spatial filter are applied to the search region. Then,
additional disc constraints are applied to identify longest chain
in top and bottom images. Candidate discs must extend onto at least
two adjacent slices. Objects at the boundary or touching the
boundary of the search region are excluded. Different threshold
values, and candidate discs' constraints are applied to the top,
and bottom half image.
[0049] In FIG. 2, the automated disc detection and numbering
algorithm 100 is a multi-step iterative process. DICOM (i.e.,
Digital Imaging and Communications in Medicine) ASSIST images of
the cervicothoracic (top half), and thorarolumbar (bottom half)
spine are first input into MATLAB 7 (The Math Works, Inc., Natick,
Mass.) for digital image processing.
[0050] Initially, these two data sets are processed to obtain
putative discs separately, utilizing different threshold values and
disc constraint parameters (block 104), with resulting upper and
lower images 106, 108 depicted in FIGS. 5 and 6 respectively. Image
volumes 106, 108 are enhanced with a tophat filter and background
noise is suppressed. A posterior edge 110 of the back is then
identified in each image 106, 108 and search regions 112, 114
assigned respectively. The algorithm 100 thresholds and applies a
median spatial filter to the search regions. Voxels 116 exceeding
threshold values are then subjected to additional constraints.
[0051] Acceptable voxels 120 must extend onto at least two adjacent
sagittal sections but not touch the boundary of the search region
112, 114. The acceptable voxels 116 must lie 6-80 mm in the
cervicothoracic (C-T), and 8-100 mm in the thoracolumbar (T-L)
region from adjacent candidate voxels 120. The centroids of these
voxel clusters 120 (candidate discs) are then connected by curved
line 122. The angle subtended by the line 122 connecting the
centroid of adjacent candidate discs 120 and the major axis of
these discs 122 must be between 45.degree. and 135.degree. in both
the C-T and T-L spine. In FIG. 3, projection analysis is used to
constrain disc assignments. Furthermore, for a disc (k) to be
considered part of a chain, its closest superior neighbor must have
k as its closest inferior neighbor and k's closest inferior
neighbor must have k as its closest superior neighbor. In FIG. 4,
an angle relative to the major axis is evaluated to constrain disc
assignments. The algorithm selects the longest disc chain (white
discs connected by curved line) in the C-T and T-L regions
respectively (FIGS. 5, 6).
[0052] In block 130, the technologist is instructed to approximate
(click on) the centroid of C2-3 at anytime during or before the
aforementioned candidate disc evaluation. The centroid of C2-3 and
the longest disc chains in the C-T and T-L spines are connected
with a straight line. Using 3D linear interpolation, the program
searches along this line and twelve adjacent parallel lines 140,
represented by only four in FIG. 7 for clarity due to their
superimposition in the sagittal projection. After applying Gaussian
filters, the algorithm 100 finds local intensity maxima along each
path. Points 142 that are less than 7.5 mm apart are grouped into
clusters. These clusters are analyzed based on orientation,
eccentricity, and spatial relationship relative to other clusters
(block 144 of FIG. 4).
[0053] Continuing with FIG. 2, if twenty-three (23) discs are
selected in block 146, the computer auto-labels theses discs or
adjacent vertebrae and stops (block 148), as depicted in FIG. 8.
Otherwise, search criteria and thresholds are refined based on
estimated inter-disc height (h) for each disc level (L) using the
following formula: (equation 1) h=0.6M for L=1, 2, or 3 and
h=M+(L-1.2)*0.05M for L>3. where M (mean)=distance along line
thru centroids/23.
[0054] In block 150, if 23 discs are not yet identified in block
146, the program 100 extends the search line inferiorly based on
the estimated position (E.sub.x,y) of the missing disc(s). E ( x j
, y j ) = ( x j - 1 , y j - 1 ) + h a .function. ( x j - 1 , y j -
1 ) .times. 1 j - 1 .times. h s .function. ( x , y ) h a .function.
( x , y ) ( Equation .times. .times. 2 ) ##EQU1## where
.differential..sub.a is the average vertebral height from a 22
subject training set, .differential..sub.s is the vertebral height
from the subject in question. The program searches for local maxima
along this line extension and 24 adjacent parallel lines. A further
determination is made in block 152 as to whether 23 discs have been
found. Iteration continues as long as twenty-three (23) discs are
not selected, as represented by block 154 that extends the search
and the further determination of block 156, which labels the
vertebrae is 23 are selected (block 157).
[0055] If not 23 discs in block 156, then in block 158 a further
determination is then made that twenty-two (22) discs are selected.
If so, the algorithm 100 will determine in block 160 whether the
last identified level (L4-5) satisfies criteria for the terminal
pre-sacral interspace suggesting a congenital variant with 23
rather than the typical 24 mobile presacral vertebrae (block 162).
To be considered a terminal disc, the centroid of the 22nd disc
must lie within 25% of its expected location (Ex,y) relative to the
21st disc. Additionally, the centroid must lie posterior to the
centroid of the 21.sup.st disc and the slope of the 22.sup.nd disc
must be positive and greater than that of the 21.sup.st disc.
[0056] If found in block 166, the discs are labeled in block 168.
Else, if the terminal disc criteria are not met in block 166, the
position of the 23.sup.rd (L5-S1) disc is estimated using Equation
2 (block 164), and search constraints refined. If the 23.sup.rd
disc is still not identified in block 166, the disc is presumed to
be severely degenerated or post-surgical and the estimated position
for L5-S1 will be accepted in block 170 and the discs thus labeled
(block 172).
[0057] If less than 22 discs are identified by the algorithm in
block 158, then the technologist will be instructed in block 174 to
approximate (click on) the centroid of the last disc. The "combine
data" step from block 144 is repeated and if necessary the "search
for additional discs" step as well. If twenty-three (23) discs are
selected in block 176, then the discs are labeled in block 178.
Else, if twenty three (23) discs are still not selected in block
176, the algorithm prints out, "Sorry, computer unable to detect
all the discs. Require technologist revision." (block 180)
[0058] The algorithm was run on an INTEL (San Jose, Calif.)
personal computer with a 2.8 Ghz Xeon processor. Computer
auto-labeling was compared to independent neuroradiologist
assignments for each patient's study. The automated spine MRI
sequencing provided a robust survey of the entire C-T-L spine in 42
seconds total acquisition time. In all patients (50/50), the
neuroradiologist could readily visualize and definitively number
all C-T-L levels on the ASSIST localizers. These included six
patients with a congenital shift in their lumbar-sacral transition;
three with 23 mobile pre-sacral vertebrae (FIG. 5) and three with
25 mobile pre-sacral vertebrae (FIG. 6). Several patients had
post-operative spines to include metallic instrumentation (FIG.
7).
[0059] Automated disc detection/numbering: the initial algorithm
tested on the first 27 surface-coil corrected studies was accurate
in 26/27 cases (96%), the single error related to a severely
collapsed vertebra (FIG. 12). The modified algorithm was accurate
in all 50/50 cases (100%) of the patients to include the original
27 patients plus 23 subsequent cases, despite the presence of
congenital variations (FIGS. 5, 6), post-operative changes (FIG.
11), and prominent disc or vertebral pathology (FIG. 13) in this
patient population. None of the 50 cases required technologist
input of more than a single interspace (C2-3) though the algorithm
provides such an iterative pathway if necessary. In the vast
majority of cases, the algorithm 100, running on a personal
computer with a 2.8 GHz processor, took less than 2 minutes to
identify and label all intervertebral discs or vertebrae.
[0060] Although the ASSIST algorithm 100 was successful in all 50
patients studied, the 7 section sagittal acquisition would be
expected to fail in subjects with severe scoliosis due to
insufficient spine coverage. As such, if significant scoliosis is
suspected, more sagittal sections could be auto-prescribed, the
cost being a proportionate increase in scan time. The automated
disc detection/numbering algorithm 100 was designed to accept any
number of sagittal sections, however, its accuracy in patients with
severe scoliosis is unknown and parameter modifications might be
required. Additionally, ASSIST algorithm 100 was designed and
tested only on an adult population. Consequently, the algorithm
would likely require additional testing and modifications to
perform optimally in the pediatric population.
[0061] While the illustrative disc detection algorithm 100
presently requires manual input of the most cephalad disc, C2-3, to
achieve maximal accuracy, it should be appreciated that automated
computer detection of this interspace may be implemented. The C2-3
disc may be readily discerned on midline sagittal head images with
a 22-24 cm FOV (Weiss 2003, 2004) or ASSIST 35 cm FOV
cervico-thoracic spine images based on several unique features.
These include the characteristic shape of the C2 vertebrae and
relatively constant relationship to the skull base and
cervico-medullary junction.
[0062] Although the illustrative version is described for MRI, it
should be appreciated that the ASSIST algorithm 100 has
applications to other imaging platforms, such as CT, substituting
automated sagittal CT spine reconstructions for the direct sagittal
MRI acquisitions facilitating automated temporal comparisons,
multi-modal multiparametric spinal analysis, and optimized
intervention. As disclosed for MRI and CT of the brain in the
cross-referenced application, direct scanner integration and
related algorithms for computer assisted diagnosis could eventually
enable "real-time" automated spine image analysis and iterative
scan prescriptions.
[0063] For example, optimally angled axial oblique sequencing could
be auto-prescribed through all interspaces or those discs
demonstrating abnormal morphology or signal characteristics on the
ASSIST or subsequent sagittal sequences. By streamlining and
converting Matlab code to C++, algorithm processing time might be
significantly shortened. Coupled with an integrated head and spine
array coil, rapid computer automated iterative prescription and
analysis of the entire neuro-axis may be possible.
[0064] In conclusion, the entire spine can be effectively surveyed
with sub-millimeter in-plane resolution MR in less than 1 minute.
All C-T-L vertebrae and discs can be readily identified and
definitively numbered by visual inspection or semi-automated
computer algorithm. We advocate ASSIST for all thoracic and lumbar
spine MRI studies. This rapid technique should facilitate accurate
vertebral numbering, improve patient care, and reduce the risk of
surgical misadventure.
Integrated Multimodality, Multi-Functional Spatial Reference and
Skin/Surface Marking System.
[0065] With internal structures labeled, there are a number of
advantages to providing an external skin/surface marking system.
There are three major configurations of the device as follows: (1)
a point localizer, (2) cross-shaped localizer grid, and (3) full
planar localizer grid/phantom.
[0066] In one version, the point localizer 500 of FIGS. 14A-14I is
a multimodality visible and compatible affixed to an adhesive
fabric strip with corresponding markings so that after application
and imaging the localizer can be removed with skin marking
remaining. The localizer can also directly adhere to the skin.
Alternatively, an ink or dye could be added to the
adhesive/undersurface of the localizer to directly mark/imprint the
skin obviating the fabric strip. For MRI and CT a small loop of
tubing could be filled with a radioattenuative solution (e.g.
containing iodine) doped with a paramagnetic relaxant (e.g. CuS04,
MgS04, Gd-DTPA). Alternatively, the tubing itself may be radiopaque
for optimal CT visualization. For nuclear imaging to include planar
scintigraphy, SPECT and PET, a port would be included to allow
filling with the appropriate radionuclide. While the above
localizers would be visible with planar radiography, a fine wire
circle or dot (e.g. lead, aluminum) could be utilized with this
modality given its very high spatial resolution. Other shapes and
corresponding adhesive markings could be utilized to discriminate
different foci and/or add further localizing capability.
Additionally, an activatable chemiluminescent mixture could be
added for thermography, optical imaging, light based 3D space
tracking or simply improved visualization in a darkened
environment.
[0067] In the second major configuration, a unique cross shaped
grouping of prefilled or fillable tubing is utilized as a grid for
cross sectional imaging with the number and position of tubes
imaged in the axial or sagittal planes corresponding respectively
to the slices z or y distance from the center. For planar
radiography, a flexible radiopaque ruled cross shaped grid is
employed. Both grids are removable from similarly ruled cross
shaped adhesive strips after patient application and imaging.
[0068] Lastly, a unique essentially planar grid/phantom is
described which may be of flexible construction and reversibly
affixed to an adhesive/plastic sheet with corresponding grid
pattern for skin marking and to serve as a sterile interface
between patient and reusable grid/phantom. The grid/phantom may
also be directly adherent to the skin for guided aspiration or
biopsy with the cross sectionally resolvable spatial reference in
place. A diagonally oriented prefilled or fillable tube overlies a
grid like lattice of regularly spaced tubing so that slice location
and thickness can be readily determined in the axial and sagittal
planes. Additionally, the spatial accuracy of the imaging modality
could be assessed and, if individual tubes are filled with
different solutions, multipurpose references for MR/CT, and nuclear
imaging could be achieved. Furthermore, if the tubing is surrounded
by a perflurocarbon or other uniform substance without magnetic
susceptibility, MR imaging could be improved by reducing skin/air
susceptibility and motion artifact. Additionally, the grid/phantom
could be incorporated in routinely utilized pads and binding
devices with or without skin adhesion and marking.
[0069] Returning to the Drawings, FIGS. 14A-14I depict an
illustrative version of a point localizer 500. In FIG. 14A, a loop
of prefilled tubing 510 (i.e., a tubal lumen shaped into a tubal
ring) is shown superimposed on and reversibly affixed to an
underlying medical grade fabric 511, which may double as an
adhesive bandage to both cover and mark a wound or puncture site.
The diameter of the tubular ring 510 may be 2 cm mid luminal, as
illustrated, or outer luminal, as perhaps preferable for
integration with the cross shaped grid configuration. Other sized
rings, to include in particular a 1 cm. diameter, may also have
merit. The tubal lumen should measure 2-5 mm in cross sectional
diameter. Cross sectional images through the ring will have
characteristic and quantifiable appearances depending on slice
thickness and distance from the center. A thin slice through the
loop's center, for example, would demonstrate 2 circles of luminal
diameter whose centers are separated by a distance equal to the
ring's diameter.
[0070] The tube lumen can be filled with an appropriate
concentration of an iodinated contrast agent for optimal CT
visualization doped with a paramagnetic relaxant such as
CuS04,MgS04, or GdDTPA to maximize MRI signal via preferential T1
shortening. Alternatively, the tubing itself may be optimally
radiopaque for CT, obviating the iodinated contrast. If desired,
for optical imaging, thermography, light based 3D space tracking,
or improved visibility in a darkened environment, one could add an
activatable chemiluminescent mixture whose critical reagents are
separated by a membrane readily breached by external force applied
to the ring.
[0071] A slightly redundant backing 512 is provided for the
adhesive side of the fabric 51 to facilitate peeling (FIG. 14B
arrows) and skin placement. With backing 512 removed, the unit 500
adheres to skin 513 as depicted in FIG. 14C. After imaging, the
loop 510, which has its own adhesive undersurface, may be removed
revealing an underlying fabric marking 514 as in FIG. 14D. The
upper surface of the fabric, or circumscribed area thereof, may
also have adhesive backing-like properties to facilitate detachment
of the ring 510. Once separated from the fabric, the loop 510 could
also directly adhere to the skin 513 as in FIG. 14E. Additionally,
the adhesive undersurface of the ring could contain a medical grade
dye or ink so that a corresponding imprint would be left on the
skin 513 after removal, potentially obviating the fabric
marker.
[0072] A port 515 may be directly integrated into the tubular ring
510 as in FIG. 14F, and a vacuum created within the lumen to
facilitate filling at the imaging center. This feature would be
critical for radionuclide scans and add flexibility for other
imaging studies.
[0073] To increase spatial reference capability and allow multiple
localizers to be discriminated from each other, the ring and
underlying fabric marking may be modified as in FIGS. 14G and 14H.
As illustrated, two tubular spokes 516 at right angles to each
other may be added with luminal diameter less than or equal to that
of the loop. Typically, the modified ring would be positioned on
the patient so that the spokes are aligned at 0 and 90 degrees as
perhaps determined by the scanner's alignment light. Subsequent
rings could be progressively rotated 90 degrees so that quadrants
I, II, III, and IV are sequentially subserved by the spokes. With
the simple ring included, this would provide 5 distinguishable
localizers. Moreover, if stacking of two rings is utilized, 30
(5.times.6) distinguishable localizer configurations are possible.
Suggested numbering would employ the base 5 system, assigning the
simple ring the value 0 and each modified ring the value of the
quadrant subserved.
[0074] Multiple localizers may also be dispensed on a single sheet
rather than individually packaged. FIG. 14I illustrates such a
sheet, demonstrating adhesive backing 517 and overlying fabric 511
with the simple ring (left side) and modified ring (right side)
localizers removed. Tabs 518 have been added to the fabric to
facilitate both removal of the unit from the backing and the
localizer from the fabric. Discontinuity of the backing (solid
lines 519) underlying the tabs would also simplify removal of the
fabric from--the backing and perforations through the backing
(dotted lines 520) would facilitate separation of individual units
from each other. If desired, a smaller diameter (e.g. 1 cm) ring
and associated backing albeit without tab could be placed within
the central space (21) bordered by the simple ring fabric 519.
[0075] Embodiments of a prefilled or fillable cross shaped
localizer grid 600 are illustrated in FIGS. 15A-15F. In FIG. 15A, a
modified tubular square 621 with diagonal dimensions of 2 cm and
containing 2 smaller caliber spokes 623 at right angles to each
other serves as the hub. Uniquely positioned rows of tubing (24)
radiate from each corner along the vertical and horizontal axes.
The luminal diameter of the radiating tubes is uniform and on the
order of 2 mm. except where indicated by dotted lines 625
corresponding to gradual tapering from 0 to the uniform diameter.
Depending on the distance from the central hub, with 1 or 2 rows of
tubing will be present with up to 4 tubes in each row as best
illustrated in a table of FIG. 15B. The lower row of tubes (i.e.
closest to skin) would correspond to increments of 1 cm. and the
upper row to increments of 5 cm so that a maximum distance of 24 cm
would be denoted by 2 full rows. To indicate positive distances,
the tubes are progressively ordered from left to right or down to
up with the reverse true for negative distances as illustrated in
FIGS. 15A-15B. Fractions of a centimeter would be indicated by the
diameter of a cross section through a tapered portion of tubing
divided by the fall uniform diameter.
[0076] The cross-shaped grid of tubing is reversibly affixed to a
medical grade adhesive fabric 626 with corresponding markings and
backing. The fabric 626 is illustrated in FIG. 15C with the
overlying tubing removed. The grid and associated fabric may come
as a single cross-shaped unit or as a modified square and separate
limbs which could be applied to the patient individually or in
various combinations. Modified squares could also link whole units
and/or individual limbs together to expand coverage, with 25 cm.
spacing between the center of adjacent squares. The tubing may be
flexible to allow the limbs to conform to curved body contours such
as the breast. Additionally, either the limbs could be readily
truncated at 5 cm. intervals or be available in various lengths for
optimal anatomic coverage.
[0077] To add further utility and integration with the previously
described point localizers, a modified ring may serve as the hub of
the cross-shaped grid with associated modification of the limbs as
illustrated in FIG. 15D. The orthogonal limbs would not have to
maintain a coincident relationship to the spokes as with the
modified square hub. Rather, by first placing and rotating a
calibrated ring adapter (FIG. 15E) about the modified loop, 1 to 4
limbs could be readily positioned at any desired angle relative to
the spokes. Pairs of male plugs 627 extending from the ring,
adapter would fit securely into corresponding holes 628 at each
limb base to ensure proper positioning. It is foreseen that one
would typically align the modified ring's spokes with the scanner's
axes and the ring adapter/limbs with the axes of the body part to
be studied. By noting the chosen angulation marked on the ring
adapter, optimal scanning planes might be determined prior to
imaging.
[0078] For planar radiography, a cross-shaped grid of radiopaque
(e.g. lead or aluminum) dots at 1 cm intervals interposed by 5 cm
spaced dashes (FIG. 15E) would minimize the imaging area obscured
by overlying radiopacity. The minute opacities could be reversibly
affixed by clear tape to an underlying marked adhesive fabric
similar to that illustrated in FIG. 15C. Alternatively, in FIG. 15F
similarly spaced radiopaque dots and dashes could be dispensed
reversibly affixed to a role of medical grade adhesive tape with
corresponding markings. Any length of this dually marked taped
could be applied to the patient to include a single dot as a point
localizer.
[0079] In a planar localizer grid/phantom 700, 1 cm spaced
horizontal and vertical tubes form a graph paper-like lattice as
illustrated in FIG. 16A. Tubes at 5 cm intervals (29) would have
larger luminal diameters (e.g. 3 mm) than the others (e.g. 2 mm).
Central vertical 730 and horizontal 731 tubes would have a smaller
diameter (e.g. 1 mm). Overlying the lattice at a 45 degree angle is
a slice indicator tube 732. Depending on the distance from the
horizontal or vertical axes respectively, axial or sagittal cross
sections through the grid/phantom (GP) would demonstrate the slice
indicator tube 732 uniquely positioned as it overlies a row of 1 cm
spaced tubes. FIG. 16B illustrates a representative axial slice
obtained 6 1/2 cm above the horizontal axis. Note that the cross
section of the slice indicator is positioned above and midway
between the sixth and seventh tubes to the right of the sectioned
vertical axis 730. Additionally, the thickness (t) of the image
section can be readily determined as it equals the cross-sectional
width (w) of the indicator minus the square root of 2 times the
diameter (d) of the indicator lumen,(t=w-V2'' d).
[0080] The GP may be reversibly affixed to an adhesive/plastic
sheet with a corresponding graph paper-like grid for skin marking
and to serve as a sterile interface between the patient and GP.
Perforations 733 may be placed in the sheet as shown in FIG. 16A to
allow ready separation of a cross-like ruled adhesive fabric
(similar to that illustrated in FIG. 15C), from the remaining
plastic sheet after imaging and removal of the GP.
[0081] The square GP should have outer dimensions equal to a
multiple of 10 cm (e.g. 30 cm as illustrated) to allow for simple
computation if GPs are placed side to side for expanded surface
coverage. Adapters could be provided to ensure secure and precise
positioned of adjacent GPs either in plane or at right angles to
each other. The GPs can be flexible or rigid in construction and be
utilized with or without skin adhesion and marking.
[0082] Tubes may be filled uniformly or with a variety of known
solutions having different imaging properties to serve as
multipurpose references. For the latter, the 5 cm spaced tubes and
slice indicator may be filled with the same optimized solution as
previously described, while each set of 4 intervening tubes could
be filled with different solutions in similar sequence. In this
fashion, identical series of 5 reference solutions would repeat
every 5 cm, allowing intraslice signal homogeneity to be assessed
as well. If 9 rather than 5 different solutions are desired,
sequences could instead be repeated every 10 cm. For MRI, the
central tubes may also be surrounded by an oil/lipid within a
larger lumen tube to serve as both a lipid signal reference and
allow for measurement of the fat/water spatial chemical shift.
Furthermore, if the GP tubing is surrounded by a perflurocarbon or
other substance without magnetic susceptibility, MR imaging could
be improved by reducing skin/air susceptibility artifact and
dampening motion. The GP may also be incorporated into a variety of
nonmodality specific pads (including the ubiquitous scanner table
pad(s)), binders, compression plates, biopsy grids and assorted
stereotaxic devices.
[0083] Two additional variations are now described, potentially
replacing the somewhat complex cross design (FIGS. 15A-15F) with an
extension of the basic point localizer (FIGS. 14A-14I) or
modification of the planar phantom/localizer (FIGS. 16A-16B). These
changes may further simplify and unify the proposed marking
system.
[0084] In the first instance, rather than packaging the ring
localizers in a sheet as illustrated in FIG. 14I, they could be
packaged in a strip or roll 800, regularly spaced at 5 cm or other
intervals (FIG. 17). The strip 800 with attached ring and/or cross
localizers could then serve as a linear reference of any desired
length. By placing two strips orthogonally, a cross-shaped grid is
created. Individual rings can be removed from the strip or rotated
to customize the grid as desired (FIG. 18).
[0085] In the second instance, by slightly modifying the square
design illustrated in FIGS. 16A-16B, an elongated rectilinear or
cross configuration (FIG. 17A) is achieved consisting of linearly
arranged squares extending vertically and/or horizontally from the
central square. One tube in each of these squares will have a
larger diameter than the other similarly oriented tubes as
determined by the square's distance from the isocenter. For
example, the square centered 10 cm above the isocenter would have
its first tube situated to the right of midline given an increased
diameter and the square centered 20 cm above the isocenter would
have its second tube to the right of midline given an increased
diameter and so on.
[0086] Cross sectional distance from isocenter would be read by
adding the distance determined by each square's diagonally oriented
slice indicator to 10 times the numberline position of the largest
diameter tube. FIG. 8B illustrates the cross sectional appearance
of an axial section obtained 121/2 cm. above isocenter. By adding
21/2 (the slice indicator position) to 10 times 1 (the tube with
largest diameter), distance is readily determined.
[0087] Alternatively, the caliber of all tubes could be kept
constant and instead an additional diagonal indicator tube passing
through isocenter added for each elongated axis (vertical with
slope of 10 and horizontal with slope of 1/10). Cross-sectional
distance from isocenter would then be determined by looking at the
position of both the additional and original diagonal indicator
tubes in reference to the cross sectionally-created number
line.
[0088] It should also be noted that localizer grids similar to
those illustrated in FIGS. 16A-16B and 18 could be constructed of
barium (or other high x-ray attenuative material) impregnated
sheets rather than tubes if computed tomography is the primary
imaging modality desired and phantom attenuation values are not
needed. This would significantly reduce the cost of the grid,
allowing disposability and retaining 1:1 compatibility with the
multifunctional tube filled grid/phantom.
Flexible Phased Array Surface Coil With Integrated Multimodality,
Multifunctional Spatial Reference And Skin/Surface Marking
System.
[0089] It should be further noted that applications consistent with
the present invention may be modified to include a sheath for and
inclusion of a flexible array MRI surface coil. Positioned
vertically, this device could be closely applied to the entire
cervical, thoracic, and lumbar-sacral spine. Additionally, the
quantity of tubes which need to be filled in the planar
configuration to uniquely denote cross-sectional positioning, has
been substantially reduced from the original embodiment.
[0090] Phased array surface coils significantly increase signal to
noise in MRI and are commonly employed for spine imaging.
Currently, such spine coils are rigid and planar in configuration.
As such, patients can only be effectively scanned in the supine
position, lying with the back against the coil. This results in
signal drop-off in regions where the spine/back is not in close
proximity to the planar coil, particularly the lumbar and cervical
lordotic regions. The invention, described herein, would reduce the
signal drop-off and allow patients to be scanned in any position.
The prone position, for example, would facilitate interventional
spine procedures that could not be performed with the patient
supine. Patients could also be more readily scanned in flexion or
extension; or with traction or compression devices. Current surface
coils also lack an integrated spatial reference and skin marking
system. Inclusion of the proposed spatial reference and skin
marking system would facilitate multi-modality image fusion and
registration as well as the performance of diagnostic or
therapeutic spine procedures, such as biopsies, vertebroplasty, or
XRT.
[0091] A grid-localizer would be adhered to the patient's spine.
Tubing would be filled with a MRI readily-visible solution such as
water doped with CuSO4. The grid itself would typically be 10 cm
wide and 70-90 cm in length to cover the entire spine. An attached
clear plastic sheath would allow introduction of a flexible array
coil such as illustrated in FIG. 9B.
[0092] The configuration of tubing would allow unambiguous
determination of the MRI scan plane (axial or sagittal) in
reference to the patient's back/skin surface. The number of thin
caliber tubes could denote the distance from (0,0) in multiple of
10 cm as illustrated in FIG. 9B (those to the right or superior
would be positive; those appearing to the left or inferior would
denote negative distances). Alternatively, as illustrated in FIGS.
9C-9D the integer distance in centimeters from a single thin tube
to the central tube could be multiplied by ten to denote distance
from (0,0). Thus, 30 cm could be denoted by a single thin tube 3 cm
from the central tube rather than by 3 thin tubes as in FIG. 9B. As
illustrated in FIGS. 9C, 9D, an axial slice taken 8 cm superior to
(0,0), would reveal the cross sectional tubes illustrated in 6d.
The thin tube being 1 cm to the right of the central tube would
denote a vertical distance of 10 cm. The diagonal oriented tube in
cross-section, being 2 cm to the left of the central tube, would
denote a vertical distance of -2 cm. Thus, the axial plane of
section would be 10-2=8 cm above (0,0).
[0093] Using the described reference/marking system affixed to the
patient's back, a diagnostic or therapeutic procedure could be
performed under direct MR guidance. Alternatively with a
corresponding radiopaque grid affixed to patient's back, the
patient could be taken out of MRI scanner and have the procedure
done with CT or fluoroscopic guidance. In either case, procedures
could be performed by hand or with a robotic arm.
[0094] Algorithms for identifying and characterizing discs and
vertebrae. A fast rule-based spine contour extraction method has
been developed. It consists of the following steps: (1) locating
inter-vertebral disc locations; (2) finding the inter-vertebral
contour using a deformable contour model; and (3) locating the
vertebral boundary and the spine contour. This method enables
automated scan prescriptions, real-time lesion detection, and
examination tailoring.
[0095] Recent advances in MRI to include the clinical
implementation of phased array-coils and parallel sensitivity
encoded imaging offer the potential for time and cost effective
non-invasive holistic screening and detailed assessment of
neuro-axis pathology, to include stroke and back pain--both leading
causes of disability in the U.S. However, optimal patient
evaluation requires individually optimized MRI sequencing, which in
turn requires real-time analysis of increasingly complex and
multi-parametric MR data. The development and integration of an
automated system emulating/approximating detailed expert analysis
while the patient is still in the magnet would significantly
improve diagnostic imaging and medical care. Millions of MRI scans
are performed each year, approximately 65% dedicated to the
evaluation of the brain and spine. Software to synergistically
improve the prescription and analysis of such scans has tremendous
commercial potential. No such product is currently available and
would be of great interest to both large medical imaging companies
engaged in computer assisted medical imaging diagnosis.
[0096] Medical Applications--Detection and Analysis of Brain
Pathology with MRI (Acute Stroke, Intracranial Aneurysms). At UC
Medical Center, Talairach referenced axial diffusion-weighted
images (DWI), whether prescribed by a technologist or a computer,
are currently obtained following the initial roll and yaw corrected
sagittal T2 sequence. If computer image analysis of the initial DWI
sequence suggests regions of acute infarction, the basic brain
protocol would be streamlined and modified to include MR
angiography and perfusion sequencing. This would respectively
permit evaluation of the underlying vascular lesion and the
detection of potential perfusion/diffusion mismatches directing
emergent neuro-vascular intervention. Stroke is the leading cause
of disability in this country. Because the time to emergent therapy
strongly inversely correlates with morbidity and mortality, the
development and implementation of the proposed computer algorithms
could significantly improve patient outcome.
[0097] In conjunction with the Mayo Clinic, researchers at UC
Medical Center are currently studying a large population of
patients at risk for intra-cranial aneurysms using MR angiography.
One of the investigators (Dr. Weiss) has developed and implemented
co-registered white and black blood MR angiography sequences which
uniquely facilitates computer-aided diagnosis and analysis of
potential aneurysms in this population. In the proposed work, such
computer algorithms will be developed and their sensitivity will be
compared against the expert standard (3 independent
euroradiologists' assessments already in place). If computer image
analysis of such initial MRA screening sequences reveals a
potential aneurysm, dedicated phase-contrast images of the putative
aneurysm could be iteratively prescribed to better characterize the
lesion and assess flow characteristics.
[0098] Using computer flow modeling and other engineering analysis,
the investigators plan to better stratify an individual aneurysm's
risk for rupture, This could lead to more optimized patient
management as the majority of brain aneurysms do not rupture and
therapeutic intervention (coiling or clipping) carries morbidity
and mortality risks. Tobacco smoking significantly increases the
incidence ischemic brain disease as well as aneurysms and their
rupture leading to catastrophic stroke.
[0099] Detection and Analysis of Spine Pathology with MRI
(Fractures, Disc Herniations). Spine pathology is another leading
cause of disability in this country. The proposed research will
improve detection and assessment of disco-vertebral degeneration,
osteoporotic and pathologic compression fractures-all potential
underlying causes of ubiquitous back/neck pain in this country.
[0100] Using advanced MR imaging techniques, the entire spinal axis
can be interrogated in the sagittal plane in less than a minute.
With this screening data, the vertebral bodies and inter-vertebral
discs can be identified and subsequently analyzed with the software
proposed for development. Based on this initial assessment, regions
of suspected pathology to include vertebral fractures and disc
hemations, could be further interrogated with more dedicated
computer driven prescriptions to confirm and better characterize
pathology. If for example, a fracture is identified, additional
multiparametric sequencing through the involved vertebrae would be
obtained to determine whether the fracture was related to
osteoporosis or underlying malignancy.
[0101] In conjunction, with the aforementioned software to
iteratively prescribe and analyze brain MRI, the entire neuro-axis
can be effectively screened and lesions characterized in a single
time-efficient scan session. Currently, such an examination is
prohibitively lengthy and requires several imaging sessions if
lesions were to be optimally characterized. Image analysis
development for this MRI project should be synergistic with that
done for the X-ray evaluation of vertebrae in the following
section.
[0102] Automated Spine MRI for Rapid Osteoporosis Screening. With
the spine labeled and imaged, further analysis is then enabled for
diagnosing conditions of the spine. Novel MRI technique provides
efficient screening and iterative assessment of patients at risk
for osteoporotic spine fractures. In particular, refinement of the
above-described Automated Spine Survey Iterative Scan Technique
(ASSIST) to optimize sub-minute morphologic screening of entire
spine with MRI; (2) to combine technique with investigational
3-point Dixon methodology to provide quantitative assessment of
vertebral marrow fat fraction (F %) and cancellous bone-induced
intravoxel spin dephasing rate (R2*); and (3) to perform
multi-variate analysis to model vertebral fracture risk as
approximated by #1, with F %, R2*, and spinal dual-energy x-ray
absorptiometric (DEXA) bane mineral density (BMD).
[0103] Osteoporosis is a disease characterized by low bone mineral
density and abnormal bone microarchitecture. Currently, it affects
about 30% of post-menopausal women, with more than 50% at risk.
With our population rapidly aging, the prevalence of osteoporosis
continues to rise. As osteoporotic-related fractures result in
major morbidity, health care expenditures, and mortality in the
elderly, this proposal addresses the DDF's desire to promote
research in Aging-Geriatrics. Moreover, by applying cutting-edge
investigational technology to this critical health-care problem,
the study fulfills Translational Research Initiative goals as
well.
[0104] The traditional criterion for assessing fracture risk is
bone mineral density (BMD) as may be measured by single-photon
absorptiometry (SPA), quantitative computed tomography (QCT),
single-energy x-ray absorptiometry (SXA), and most commonly
dual-energy x-ray absorptiometry (DEXA). Unfortunately, while
negatively correlated with fracture risk, BMD by itself remains an
unsatisfactory predictor. Consequently, investigational work has
increasingly focused on ultrasound and MRI. The latter technique
has the unique potential to quantify fractures, which are highly
correlated with subsequent risk of fracture; differentiate between
osteoporosis and other underlying pathology, such as metastases;
and target therapy such as vertebroplasty. MRI researchers have
also demonstrated improved fracture risk prediction by combining
DEXA measurements with Dixon sequence derived F % (positively
correlated) and R2* (negatively correlated). Unfortunately, MRI of
the spine has been too time intensive and costly to justify as an
osteoporosis-screening instrument.
[0105] To rectify this important shortcoming, integration is made
of the 3-point Dixon technique with our novel automated sub-minute
sub-millimeter resolution total spine screen. This should afford
rapid high-resolution morphometric assessment, as well as, the
separation and quantification of fat, water, and R2*. We plan to
test this methodology on 50 post-menopausal women who have been
referred for a DEXA scan.
[0106] In particular, a novel MRI technique improves current
screening, assessment, and surveillance of the elderly at risk for
osteoporotic spine fractures. As osteoporotic fractures result in
major morbidity, health care expenditures, and mortality in the
geriatric population, this proposal directly addresses the Dean's
Discovery Fund's desire to promote research in Aging/Geriatrics.
Moreover, by applying cutting-edge investigational technology to
this critical health-care problem, the study fulfills Translational
Research Initiative goals as well.
[0107] Osteoporosis and related fractures are a leading cause of
morbidity, disability, decreased quality of life and mortality in
the aged. (2-4) The wide range of therapeutic options available for
prevention and treatment require effective screening, assessment,
and monitoring of geriatric patients at risk for osteoporotic
fractures. Conventional measurements including bone mineral density
(BMD) analysis are imperfect predictors of fractures. (4)
MRI-derived parameters hold promise for improved risk prediction
and fracture evaluation. (5) Unfortunately, MRI has been too time
intensive and costly to justify as an osteoporosis-screening
instrument. To rectify this important shortcoming, we propose
refinement and integration of MRI derived parameters of bone
quality with our novel rapid high-resolution total spine screen (1)
The long term goal is to promote geriatric patient care by
providing improved risk assessment, identification and
characterization of fractures.
[0108] The central hypothesis is that our computer automated MRI
technique will efficiently screen and assess elderly
post-menopausal women at risk for osteoporotic spine fractures.
More specifically, our multi-parametric approach will 1) improve
fracture risk prediction and 2) accurately identify and
characterize existing fractures. The following specific aims will
be pursued to test this hypothesis:
[0109] Improve vertebral fracture risk prediction: MRI derived
measurements of vertebral morphology and bone quality (fat
fraction, F %; transverse relaxation rate, R2*) will be calculated
for each vertebra. These parameters will be analyzed in conjunction
with standard dual-energy x-ray absorptiometric (DXA) to develop a
model for prediction of vertebral fracture risk.
[0110] Accurately identify and characterize existing vertebral
fractures: Our novel MRI Automated Spine Survey Iterative Scan
Technique (ASSIST) will be adapted to provide automated morphologic
assessment of vertebrae and detect vertebral fracture deformities.
This functionality will be compared to lateral thoracolumbar x-ray,
which currently serves as the clinical standard.
[0111] Osteoporosis is an important geriatric health issue and
poses a most serious public health problem. With life expectancies
increasing, the financial and human costs associated with
osteoprotic fractures will multiply exponentially throughout the
world. Vertebral fractures are strongly correlated with age (mean
65 years) but even more so with menopause. In the United States,
one out of two women and one in four men over age 50 will have an
osteoporosis-related fracture.
[0112] Osteoporosis is a metabolic disease characterized by low
bone mineral density and abnormal bone microarchitecture increasing
fracture risk. The traditional criterion for assessing fracture
risk is bone mineral density (BMD) as may be measured by
single-photon absorptiometry, quantitative computed tomography,
single-energy x-ray absorptiometry, or most commonly dual-energy
x-ray absorptiometry (DXA). Unfortunately, while negatively
correlated with fracture risk, BMD by itself is not a perfect
predictor.
[0113] Fractures are the ultimate manifestation of lost bone
structural integrity. One fractured vertebra increases the risk of
subsequent vertebral fracture 5-fold. Consequently, low resolution
morphometric x-ray absorptiometry and/or more precise high
resolution conventional thoracolumbar spine x-rays are often
ordered to supplement BMD measures. Vertebral fractures are
commonly assessed on x-rays semiquantitatively and reported using
Genant's 0-3 grading scale with grade 1 corresponding to a fracture
deformity of 20-25%. More mild deformities are typically not scored
but are also somewhat associated with lower BMD and increased
fracture susceptibility.
[0114] Conventional radiography, however, has several shortcomings
in the detection of vertebral fractures. Different technical
parameters in the acquisition of lateral spine radiographs
influence vertebral dimension measurements, thus degrading
reproducibility, especially in scoliotic patients. Vertebral body
outlines are more difficult to visualize on x-ray in the population
most at risk for osteoporotic fracture due to their reduced BMD.
Radiographs also have limited ability to assess fracture
chronicity, etiology, or potential effectiveness of targeted
intervention to include vertebroplasty or kyphoplasty. Moreover,
ionizing radiation from serial radiographic examinations must be
taken into account, especially when considering clinical trials
typically requiring numerous exposures.
[0115] Consequently, investigations have increasingly focused on
ultrasound and MRI. MRI has the unique potential to simultaneously
quantify fractures; differentiate between osteoporosis and other
underlying pathology, such as metastases; and appropriately target
therapy such as vertebroplasty. Additionally, spinal MRI affords
direct visualization of the intervertebral discs, spinal canal,
bone marrow, and neural tissue. MRI can accurately quantify
vertebral morphology. Cyteval and colleagues have recently
demonstrated the accuracy and reproducibility of MRI for the
determination of vertebral body dimensions. They also found that
the sagittal midline area was highly correlated with whole
vertebral body volume and that each vertebra was proportional to
other vertebrae in the same individual.
[0116] MRI can also provide bone quality measurements related to
osteoporosis. MRI researchers have demonstrated improved fracture
risk prediction by combining DXA measurements with Dixon sequence
derived fat percentage--F% (positively correlated) and transverse
relaxation rate--R2* (negatively correlated). When trabeculae are
not aligned with the magnetic field, susceptibility differences at
the interface between trabecular bone and bone marrow increase R2*.
Consequently, R2* measurements reflect both trabecular bone
structure and orientation.
[0117] The Dixon technique exploits the resonant frequency
differences between fat and water to separate the water signal
intensity from the fat signal intensity. This frequency difference
is measured as a phase difference in the acquired data. Acquisition
of three separate measurements or Dixon echoes allows generation of
a water image, a fat image, and a magnetic susceptibility map from
which F % and R2* can be derived. (6)
[0118] To date, while promising, MRI examinations have been too
time intensive and costly to justify for osteoporosis-screening. To
rectify this important shortcoming, we propose integration of the
three-point Dixon technique with our novel automated sub-minute
sub-millimeter resolution total spine screen. This should afford
rapid high-resolution assessment of both vertebral morphometry and
bone quality.
[0119] Rapid Osteoporosis Screening: Sagittal FGRE (TR 58; TE 2.1;
30.degree. flip; 7 slices; 4 mm skip 1 mm; 35 cm FOV.times.2=70 cm)
with contiguous superior and inferior stations. The imaging
parameters have been selected to emphasize contrast between
vertebral discs and bodies with full coverage from the cervical
spine through the sacrum in 42 seconds.
[0120] Three-Point Dixon Technique: Sagittal FSE Dixon (7 slices; 4
mm skip 1 mm; 44 cm FOV) covering T4-S1 and prescribed using the
superior ASSIST station as localizer. The Dixon technique exploits
molecular resonant frequency differences between fat and water to
produce high resolution fat, water, and transverse relaxation rate
(R2*) images.
[0121] While the present invention has been illustrated by
description of several embodiments and while the illustrative
embodiments have been described in considerable detail, it is not
the intention of the applicant to restrict or in any way limit the
scope of the appended claims to such detail. Additional advantages
and modifications may readily appear to those skilled in the
art.
* * * * *