 |
Description  |
|
|
BACKGROUND AND DISCUSSION OF THE INVENTION
Diagnostic techniques that allow the practicing clinician to obtain high
fidelity views of the anatomical structure of a human body have proved
helpful to both the patient and the doctor. Imaging systems providing
cross-sectional views such as computed tomographic (CT) x-ray imagers or
nuclear magnetic resonance (NMR) machines have provided the ability to
improve visualization of the anatomical structure of the human body
without surgery or other invasive techniques. The patient can be subjected
to scanning techniques of such imaging systems, and the patient's
anatomical structure can be reproduced in a form for evaluation by a
trained doctor.
The doctor sufficiently experienced in these techniques can evaluate the
images of the patient's anatomy and determine if there are any
abnormalities present. An abnormality in the form of a tumor appears on
the image as a shape that has a discernible contrast with the surrounding
area. The difference in contrast is due to the tumor having different
imaging properties than the surrounding body tissue. Moreover, the
contrasting shape that represents the tumor appears at a location on the
image where such a shape would not normally appear with regard to a
similar image of a healthy person.
Once a tumor has been identified, several methods of treatment are utilized
to remove or destroy the tumor including chemotherapy, radiation therapy
and surgery. When chemotherapy is chosen drugs are introduced into the
patient's body to destroy the tumor. During the course of treatment,
imagers are commonly used to follow the progress of treatment by
subjecting the patient to periodic scans and comparing the images taken
over the course of the treatment to ascertain any changes in the tumor
configurations.
In radiation therapy, the images of the tumor generated by the imager are
used by a radiologist to adjust the irradiating device and to direct
radiation solely at the tumor while minimizing or eliminating adverse
effects to surrounding healthy tissue. During the course of the radiation
treatment, the imaging system is also used to follow the progress of the
patient in the same manner described above with respect to chemotherapy.
When surgery is used to remove a tumor, the images of the tumor in the
patient can guide the surgeon during the operation. By reviewing the
images prior to surgery, the surgeon can decide the best strategy for
reaching and excising the tumor. After surgery has been performed, further
scanning is utilized to evaluate the success of the surgery and the
subsequent progress of the patient.
A problem associated with the scanning techniques mentioned above is the
inability to select and compare accurately the cross section of the same
anatomical area in images that have been obtained by imagers at different
times or by images obtained essentially at the same time using different
image modalities, e.g., CT and MRI. The inaccuracy in image comparison can
be better appreciated from an explanation of the scanning techniques and
how the imaging systems generate the images within a cross-sectional
"slice" of the patient's anatomy. A slice depicts elemental volumes within
the cross-section of the patient's anatomy that is exposed or excited by a
radiation beam or a magnetic field and the information is recorded on a
film or other tangible medium. Since the images are created from slices
defined by the relative position of the patient with respect to the
imager, a change of the orientation of the patient results in different
elemental volumes being introduced into the slice. Thus, for comparison
purposes, two sets of image slices of approximately the same anatomical
mass taken at different times do not provide comparable information that
can be accurately used to determine the changes that occurred between two
image slices in the sets, since it is unknown to what extent the two
individual image slices selected from the respective sets depict identical
views.
The adverse effects on the medical practice of such errors is exemplified
by diagnostic techniques utilized by the surgeon or others in diagnosing a
tumor within a patient. If a patient has a tumor, its size density and
location can be determined with the help of images generated by a scanning
device. For the clinician to make an assessment of the patient's
treatment, two scanning examinations are required. The patient is
subjected to an initial scan that generates a number of slices through the
portion of the anatomy, for instance the brain, to be diagnosed. During
scanning, the patient is held in a substantially fixed position with
respect to the imager. Each slice of a particular scan is taken at a
predetermined distance from the previous slice and parallel thereto. Using
the images of the slices, the doctor can evaluate the tumor. If, however,
the doctor wants to assess changes in the configuration of the tumor over
a given period of time, a second or "follow-up" scan has to be taken.
The scanning procedure is repeated, but since the patient may be in a
position different from that in the original scan, comparison of the scans
is hampered. Slices obtained at the follow-up examination may be
inadvertently taken at an angle when compared to the original slices.
Accordingly the image created may depict a larger volume than that which
was actually depicted before. Consequently, the surgeon may get a false
impression of the size of the tumor when comparing scans taken at
different periods. Because of this, slice-by-slice comparison cannot be
performed satisfactorily.
Similarly for certain surgical techniques it is desirable to have accurate
and reliable periodic scans of identical segments of the tumor within the
cranial cavity. If the scans before and after surgery are inaccurate, the
doctor may not get the correct picture of the result of surgery. These
same inaccuracies apply to other treatments such as chemotherapy discussed
above.
Additionally, with regard to imaging systems and the integral part they
play in surgical and other tumor treatment procedures, there is a dearth
of methods currently existing that allow a determination of a desired
location within the body at a given time. For example, U.S. Pat. No.
4,583,538 to Onik, et. al. discloses a localization device that is placed
on a patient's skin which can be identified in a slice of a CT scan. A
reference point is chosen from a position on the device which exactly
correlates to a point on the CT scan. Measurements of the localization
device on the CT scan is then correlated to the device on the patient.
Exterior devices have been utilized in an attempt to solve some of these
problems with accuracy such as that shown in U.S. Pat. No. 4,341,220 to
Perry which discloses a frame that fits over the skull of a patient. The
frame has three plates, each defining a plurality of slots on three of
four sides. The slots are of varying lengths and are sequentially ordered
with respect to length. Frame coordinates defined and found on the frame
correspond to the varying heights of the slots. When slices of the skull
and brain are taken by an imaging device, the plane formed by the slice
intersects the three plates. The number of full slots in the slice are
counted with respect to each plate to determine the coordinate of a target
site with the brain. Accordingly, only one CT scan is needed to pinpoint
the coordinates of the target.
Other attempts have included the use of catheters for insertion into the
anatomy. For example, U.S. Pat. No. 4,572,198 to Codrington discloses a
catheter with a coil winding in its tip to excite or weaken the magnetic
field. The weak magnetic field is detectable by an NMR device thus
pinpointing the location of the catheter tip with respect to the NMR
device.
Applicant's invention largely overcomes many of the deficiencies noted
above with regard to imagers used heretofore. The invention relates to a
method and apparatus for insuring that scans taken at different times
produce images substantially identical to those of previous scans even if
they are from different image modalities at different times. This insures
that a more accurate assessment of any changes in anatomy is obtained. As
a result, the doctor can be more certain as to the size, location and
density of the tumor, or a section thereof, that is located in the cranial
cavity.
This ability will enhance the use of surgical techniques in removing or
otherwise eliminating the tumor in particular by those noninvasive
techniques such as laser technology. By having the ability to define
accurately the tumor location and size, laser beams can be focused
directly on the tumor. Intermittently, as part of surgical techniques,
scans can be made to determine if the tumor has moved or substantially
changed in size as a result of the surgery. The laser or other surgical
instrument can be adjusted accordingly. Because of the accuracy of the
imaging techniques produced by the invention, the doctor can be confident
that the amount of healthy tissue destroyed during surgery is minimized.
A method adopted by the invention disclosed herein utilizes fiducial
implants or implants to define a plane which cooperates with the imager,
or other computer, and particularly the data processing capabilities of
the imager to insure that subsequent scanning results in slices
substantially parallel to those taken during the initial scan. The
fiducial implants are implanted beneath the skin into the calvaria and are
spaced sufficiently from one another to define a plane. The patient with
these implants implanted is placed in the scanning device in the
conventional manner and scanned to provide the images of consecutive
parallel slices of a given thickness along a predetermined path through
the cranial cavity.
As the scans are taken, one or more slices will be needed to accommodate
part or all of each fiducial implant. The computational features of the
imager or other computer will take into account the spatial relationship
between any selected plane of a slice and that plane defined by the
fiducial implants. Because of this capability, images taken in subsequent
scans at different points in time, at different angles can be
reconstructed to be substantially identical with the slices taken
originally.
Fiducial implants for this purpose are specially configured and made of
material that enables their implantation into the skull and the ability to
be detected by scanning devices. The fiducial implant as disclosed herein
is configured to insure that during implantation it does not have adverse
effects on the skull such as cracking or extending through to the cranial
cavity. Nor is it sufficiently exposed between the skull and the skin to
distort any external features of the anatomy. Furthermore, the fiducial
implant is positioned at least on a portion of the skull at the interface
of the skin and the bone of the skull to facilitate its imaging by the
imager. At least a portion of the implant is symmetrical in cross-section
such that slices taken of the cranial cavity for example can be used to
locate the center of mass of the implant. This insures accuracy in using
the implant image as a reference point to transform the subsequent slices
of the follow-up examination into the proper position and orientation.
The above has been a description of certain deficiencies in the prior art
and advantages of the invention. Other advantages may be perceived from
the detailed description of the preferred embodiment which follows.
BRIEF DESCRIPTION OF THE DRAWINGS
A more complete appreciation of the present invention and many of the
attendant advantages thereof will be readily obtained, as the same becomes
better understood by reference to the following detailed description, when
considered in connection with the accompanying drawings, wherein:
FIGS. 1a, 1b, and 1c show side and overhead views of fiducial implants.
FIGS. 2a and 2b show a side and overhead view of a preferred positioning
scheme of fiducial implants in the skull.
FIG. 3 is an offset view of two coordinate systems that have undergone
translation with respect to each other.
FIG. 4 is an offset view of two coordinate systems that have undergone
rotation with respect to each other.
FIG. 5 and FIGS. 5a, 5b and 5c are offset views of two coordinate systems
that have undergone translation and rotation with respect to each other.
FIG. 6 is a flow chart with respect to determining the same point P at two
different times in an internal coordinate system to the body.
FIG. 7 is a side view of a preferred embodiment of the present invention.
FIG. 7a is an enlarged view of a portion of FIG. 7.
FIG. 8 is a flow chart with respect to determining the location of a point
P in an internal coordinate system with respect to an external coordinate
system.
DESCRIPTION OF THE PREFERRED EMBODIMENT
In FIG. 1, there is shown a fiducial implant 10 for the human body that is
detectable by an imaging system. The fiducial implant comprises a first
portion 12 and a second portion 14. The first portion 12 is configured to
be detected by an imaging system (when placed beneath the skin.) The
second portion 14 is configured for fixed attachment to the bone beneath
the skin without penetrating entirely through the bone and without
fracturing the bone, as will be described in more detail later. The first
portion 12 is of detectable size and comprised of a material for detection
by an imaging system and sufficiently small to provide minimal distortion
of the skin when placed at an interface between the skin and the bone as
will be described in more detail later. First portion 12 also has at least
a portion which is spherical and defines a surface for cooperating with a
tool for securing the second portion 14 to the bone. Additionally, the
placement of three fiducial implants 10 into a portion of anatomy of the
human body allows for the recreation of a particular image slice of the
portion of the anatomy taken by an imaging system in order to duplicate
images taken at the first time period, that is, at the initial
examination. This provides a doctor with the ability to accurately follow
the progress of treatment on selected slices representing the anatomy of
interest.
Moreover, the existence of three fiducial implants 10 allows a target (a
tumor for instance) to be identified relative to an external coordinate
system. The portion of anatomy with the target may then be operated on,
for instance, robotically, or precisely irradiated.
To allow for the accurate comparison of image slices from at least two
distinct periods of time, the three fiducial implants 10 are first
implanted into a body of a patient at a desired region of interest. The
patient is then placed in an imaging system and images of a series of
cross-sectional slices are obtained that include, for example, the volume
of the tumor which is the primary target of interest. From the imaging
data obtained, the three fiducial implants are located and an internal
coordinate system is defined with respect to them. If it is so desired,
the image data may be further reformatted to show image slices whose
direction is different from that obtained originally during the imaging
period. Depending on the diagnostic information that these image slices
reveal, appropriate decisions with regard to surgery, chemotherapy or
radiation therapy on a patient may be made. The imaging data can also be
used from several different types of images, such as CT, PET or NMR to
obtain the same view of the anatomy but with different qualities stressed.
If it is decided to obtain further imaging data at a later time, then the
patient is returned to the imaging system and the procedure for obtaining
image data is repeated. The fiducial implants 10 are located with respect
to the second imaging session and the same internal coordinate system is
defined relative to the implants 10. Once the same internal coordinate
system is defined with respect to the second imaging session, the
translation and rotation of the internal coordinate system and the images
with it is determined with respect to the coordinate system established at
the first imaging session. An image slice identified from the first
imaging session that is to be used for diagnosis, is recovered from the
second imaging session. The two image slices, one from the first image
session and one from the second image session, are then compared to
determine what changes, if any, have occurred in the anatomy of the
patient.
The following discussion relates to the three-dimensional non-collinear
coordinate system defined by three distinct non-collinear points. In
referring to such a system, it is understood that a three-dimensional
coordinate system requires three non-coplanar vectors. This is a course
achieved with three non-coplanar points. A vector orthogonal to aa plane
defined by two vectors (thee non-collinear points) can be obtained through
the cross product of the planar vectors. It should be understood that the
vectors are normalized.
More specifically, a 3-dimensional noncollinear coordinate system requires
three distinct noncollinear points to be fully defined. If there are more
than three identifiable points, the system is over-determined and three
points have to be chosen to define the coordinate system. If there are
less than three identifiable distinct points, the system is undetermined
and a position relative to the one or two identifiable points will not be
defined.
The known location of three distinct points identifies a plane upon which
an orthogonal coordinate system can be established. If the three points
are fixed in place relative to each other over time in the body, a
coordinate system can be established that is also fixed in time. The
ability to define a fixed internal coordinate system to the human body
over time has important ramifications. A fully defined internal coordinate
system that is fixed in place over time with respect to some location in
the body permits comparison of subsequent images of the body taken into
imaging systems such as CT scans, NMR scans or PET scans to name a few.
More precisely, these comparisons will allow a diagnostician to see what
change, if any, has occurred within the body at a predetermined location.
By utilizing a fixed coordinate system relative to the body, the same
coordinates can be compared over time. However, the tissue or body
material is not necessarily fixed in place relative to a predetermined set
of coordinates over time. After the passage of time, the tissue may have
shifted, a change not uncommon following surgery. Nevertheless, the
ability to compare various properties (depending on the type of images) of
the tissue at the same coordinates and at different times is a great
advantage for diagnostic purposes.
In principle, the three points (that are necessary) to define a coordinate
system can be chosen in a variety of ways. In one embodiment with respect
to the brain or head region, the two ears and a tooth, or the two ears and
the nose may comprise the three points. Alternatively, an image slice of
the skull could provide a set of points from which the three points would
be chosen to create the coordinate system for the body. Preferably, three
fiducial points that are implanted into the body, and create high contrast
images during scanning, provide the most reliable way to define a
coordinate system. Ideally the three points should be in the same
approximate area of the body that is under analysis, and also should be
identifiable and measurable by different imagery systems, such as CT
imagers and NMR imagers.
To create a fully defined coordinate system the detection of three distinct
noncollinear fiducial points is required. With respect to creating a
fully-defined coordinate system anchored to the human body, the
requirement of detection dictates the need that fiducial implants 10 are
made of a material that is detectable by a system imaging the human body.
The fiducial implant 10 has a first portion 12 that provides means for
marking a predetermined position within a body. See FIG. 1. First portion,
or marker 12, ideally provides a high contrast in an image compared to the
surrounding material. The material marker 12 is made of also provides as
little distortion as possible to the image so the appearance of artifacts
is kept to a minimum. Marker 12 is also safe for use in the human body and
is unobtrusive so no discomfort or self-consciousness is experienced by a
wearer.
Marker 12 exhibits symmetrical integrity to facilitate its location by the
imaging system. When marker 12 is scanned, the symmetry insures that any
plane through the implant provides essentially the same image and the
ability to locate its center of mass. The importance of being able to
identify the center of the marker 12 lies in the fact that the same exact
point can be reproductibly found for use in defining the coordinate
system. Error is thus minimized from subsequent recreations of the same
coordinate system due to displacement of the coordinate system from a
previous alignment. For instance, a sphere is the ideal shape for a marker
12 with respect to symmetrical integrity since the image of any plane of
the sphere is always a circle.
By knowing the radius of the spherical object and applying standard
algorithms, the center can be determined of the spherical marker 12 from
any plane passing through the sphere. The algorithm for determining the
center of a sphere may require operator interaction to mark the
approximate location of the implant. The center of mass can be determined
with successful approximation from the boundary of the circular profile
identified through the operator's interaction. For instance, by having
information about the density of the fiducial implant's image and assuming
it spherical, then scan profiles through its image result in bell-shaped
distributions, the boundary points of which can be determined therefrom.
From the boundary points the center of mass is computed. This may require
additional slices depending on the size of the fiducial implant and its
relative position with respect to adjacent slices, particularly when the
physical size of the implant exceeds that of the scan slice.
When the centers of mass of the 3 fiducials (10a, 10b, 10c) are determined,
then two of them (10 a, 10b) define for instance the x-axis vector of the
coordinate system and the vector cross product of vectors 10a, 10b and
10a, 10c fully determine the coordinate system as shown in FIG. 5a which
is described more fully below.
Marker 12, which is 1 to 10 and preferably 4 millimeters in diameter, can
be made of, for example, titanium in the form of a hollow sphere. The
hollow of the sphere can be, for example, filled with agarose gel having
various desired dopants, the choice of which depends on the imaging system
used to best accent or highlight the marker 12. Marker 12 is intimately
connected to a second portion, anchor 14, of the fiducial implant 10.
The anchor 14 provides means for anchoring the marker 12 into the body. The
site of preference for anchoring the marker 12 in the body is bone, since
it provides a good material to hold the implant means in place and also
because bone stays in a fixed position over time in the body. Anchor 14 is
long enough to penetrate into the bone to which it is anchored, and long
enough to be firmly embedded without fracturing the bone. Anchor 14 is 1
to 10 and preferably 3 millimeters long. Preferably the anchor 14 should
be screwed into the bone, rather than driven with an impact tool to lessen
the chance of fracturing the bone. Anchor 14 can also, for example, be
made of titanium.
The fiducial implant 10 also has means 16 for receiving force so the anchor
14 can be fixedly secured to the body. Where anchor 14 is a screw,
preferably an indention 16 in the shape of a polygon recess to receive an
allen wrench is located in marker 12. The use of an allen wrench with the
associated polygonal recess has more symmetrical integrity than the cross
shaped receptor site for a phillips screw driver or a single groove
receptor site for a standard screw driver.
The implantation of a fiducial implant 10 having an anchor 14, in this case
a screw, preferably utilizes a trocar not shown, to penetrate the skin and
reach a desired bone site. The trocar is first placed on the skin over the
desired anchoring site and a piercing rod therein is forced through the
skin. The piercing rod within the trocar is then removed while the trocar
is kept in place. A rod with an allen wrench head fitted to the polygonal
indentation 16 in the marker 12 of the implant 10 is inserted into the
trocar until the screw 14 portion of the implant 10 contacts the anchoring
site, for instance bone. Force is then applied to the portion of the rod
extending out the trocar until the implant 10 is embedded into the bone.
Such a procedure is accomplished under local anesthesia and should only be
about 5 minutes in duration.
The placement of the three fiducial implants 10 depends on the portion of
the anatomy to be evaluated. Essentially, three fiducial implants 10 are
placed in three locations such that they are readily identifiable and the
locations are fixed with respect to each other over time. If, for example,
a study of the skull and brain is to be undertaken, preferably an implant
10A is placed on the midline of the skull 18 just above the hairline, with
the other two implants 10B, 10C being placed on the right and left side,
respectively, of the midline in a posterior position to the midline
implant 10A. See FIGS. 2a and 2b which are a frontal and overhead view of
the skull 18, respectively. Another example of an area of interest could
be the torso, with one fiducial implant 10 placed on the midline of the
sternum and the other two fiducial implants 10 placed laterally thereto on
the right and left side, respectively, and in a rib. Or, one fiducial
implant 10 can be placed in the spinous process of a vertebra in the
midline and the other two fiducial implants placed in the right and left
illiac crest, respectively.
The fiducial implants 10 can alternatively be implanted temporarily into a
patient in situations where there is only a short term need for their
presence. For instance, the method of radiation therapy that is described
herein can use either fiducial implants | | |
