 |
Description  |
|
|
BACKGROUND OF THE INVENTION
The present invention relates to the field of surgery and of medical
observation, and more particularly relates to methods and devices for
positioning a therapeutic or diagnostic tool as a function of
three-dimensional images, that can be images carried out before hand
(preoperation images) of a patient's organ.
In the present application, "tool" is to be construed as any therapeutic or
diagnostic means carried out on a patient. The tool can be, for example, a
device adequate to insert a screw in a patient's bone, a needle to carry
out a puncture or simply to guide an optical fiber, a radiation
transmission apparatus designed to act on a tumor, or a medical imaging
device such as a gamma-scintigraphy camera, a positron emission tomography
(PET) apparatus, or a magnetoencephalography (MEG) apparatus.
In other words, an object of the invention is to recover, during a surgical
procedure, the morphological information that previous three-dimensional
(3D) examinations have provided. This situation is very frequently
encountered in the medical domain. Two examples of such a situation are
given below.
1. Radiotherapy
Radiotherapy consists in projecting onto a predetermined patient's region,
or on one of his organs, a radiation beam so as to destroy or eliminate
tumors existing in these organs. Such therapy treatments must generally be
carried out periodically and repeatedly. Therefore, at each medical
intervention, the radiation source must be repositioned with respect to
the patient in order to irradiate the selected region with the highest
possible accuracy to avoid irradiating adjacent patient's organs on which
radiation beams would be harmful.
The initial diagnostic procedure has generally been carried out by an X-ray
scanner or by magnetic resonance imaging (MRI) that allow to visualize the
target site and the obstacles, and therefore to define an optimal
therapeutic procedure. The difficulty is to reposition the patient when he
must receive the therapy treatment. The method that is presently used
consists in simulating the intervention by replacing the linear
accelerator (radiation source) by an X-ray tube. Two radiographies (face
and profile) of the patient are then obtained and are visually compared
with the previous 3D information provided by the X-ray scanner or the MRI
apparatus. The patient is moved, and radiographies are made again until
the patient's positioning is deemed satisfactory. A light beam then
materializes a marking on the patient's skin, and this marking is
inscribed on the patient's skin itself with a marker. During the
irradiation session, the patient is moved until these marks on this skin
are in coincidence with light beams, that are identical for the linear
accelerator and for the simulation radiologic system. This conventional
method has numerous drawbacks. One of them is to necessitate moving the
patient on the medical table so that he takes the desired position, which
is not easy and may compel the patient to stay in an uncomfortable
position, or cause the patient to be so crisped that he will not be able
to stay in the desired position. It should be pointed out that the
patient's position must be as accurate as possible. For example, in the
case of the final radiotherapy session of the prostate, where this organ
only, included within a sphere of 4 cm in diameter, is targeted, with the
conventional methods, it is necessary to irradiate an area of 10 cm in
diameter to ensure that the prostate is appropriately reached. Of course,
this is not harmless for the adjacent organs.
2. Orthopedics
In orthopedics applications, the point is to provide an assistance for
introducing an object (for example a screw) in the human body (usually
into a bone) according to a linear path. The shape of the bone is
conventionally studied by 3D imaging (X-ray scanner or MRI) and the
procedure of the medical intervention is then defined as a function of
this image, especially as regards the exact direction along which the tool
must be inserted in order to reach exactly the desired position without
passing through regions of the body where it can be harmful.
Conventionally, during the surgical procedure, radiography is used to
control the introduction of the tool. Under usual conditions, the major
drawback is that it is not possible to simultaneously make face and
profile images both of the target and of the object to be introduced. This
method can therefore be used only a posteriori, as a means for verifying
the appropriate positioning of the object (the progression being tracked
in real time on one projection only). Additionally, such a verification
cannot be made quantitatively. The interpretation is made by the operator
alone. Present studies show that with such techniques, in approximately
10% of cases, positioning is not ideal.
More recently, it has been proposed to improve the use of a radiologic
system in an operation site to implement techniques for matching
three-dimensional (3D) with two-dimensional (2D) images (refer to S.
Lavallee et al., "Matching 3D smooth surfaces with their 2D projections
using 3D Distance Maps" SPIE 1570, pp. 322-336, 1991). The problem with
this method is that the technique is expensive because it requires the use
of a sophisticated radiologic system capable of studying the target on
several incidences and providing a signal that can be digitized. To avoid
the drawbacks of the radiologic systems, some medical teams have proposed
to make therapy interventions using X-ray scanner or MRI apparatus. It is
then easy to make comparison between the pre-operation and on-site images
that are carried out with the same means. However, such approaches have
many drawbacks due to the fact that they involve the simultaneous use of
sophisticated imaging tools and surgical tools. Among these drawbacks, can
be cited:
surgical constraints (asepsis . . . ),
long immobilization (for the duration of a surgical procedure) of expensive
materials (X-ray scanner or MRI apparatus) while it is desirable to use
such apparatuses continuously for observation tasks, for a better economy
of operation,
the need for specific surgical tools (in the case of the MRI system, these
tools must not generate artifacts and, more particularly, must be
non-ferromagnetic); this involves, for example, the use of very expensive
tools made of titanium,
the geometry of the imaging apparatuses (small-diameter channel) renders
the introduction of the surgical tools difficult.
However, the development of such 3D image matching techniques, despite all
the above-mentioned drawbacks, shows the interest of the medical teams in
any means capable of improving the accuracy of the introduction of
surgical tools.
SUMMARY OF THE INVENTION
An object of the invention is to provide a simple and inexpensive method
for making a series of on-site images of an organ that allows to position
a tool in a predetermined manner with respect to a series of 3D
pre-operation images.
To achieve this object, the present invention provides the use, for making
on-site images, of a device providing a 3D morphologic image of surface
points of the organ of interest or a skin region fixed with respect to the
organ. Then, this surface point image is combined (matched) with the
pre-operation 3D functional image that also contains information on the
localization of the surface points (of the organ or skin region).
To obtain a 3D morphologic image of the surface points of the organ, the
invention provides the use of echography probes.
To obtain a 3D morphologic image of the surface points of a skin region
fixed with respect to the organ (for example an image of a portion of the
patient's head that is fixed with respect to the brain), the invention
provides the use of optical imaging devices.
Further, the invention provides to align the coordinates of the tool with
respect to the coordinate system of the organ which is in turn defined by
the preoperation image.
A second object of the invention is the matching of an image from a second
device that provides, as indicated, a 3D morphologic image of surface
points associated with an organ, and an image from a specific apparatus
disposed within the same surgical site, such as a gamma-scintigraphy
camera, a positron emission tomography (PET) apparatus, a
magnetoencephalography (MEG) apparatus, or a synchrotron radiation
apparatus, that provides functional information on specific regions of
this organ, thus making it possible to position these specific regions of
the organ.
To achieve this second object, the invention provides for previously
localizing an initial position of the first device and of the specific
apparatus by making them pinpoint a same target that is visible by both of
them (for example in ultra-sonic frequencies and in gamma rays in the case
of an echography probe and of a gamma-scintigraphy camera).
One of the originalities of the invention lies in the use of a device that
does not provide functional information but only images of surface points
to serve as an intermediate device operable for matching different
coordinate systems. In particular, the idea of using an echography probe
to carry out this coordinate alignment is one of the aspects of the
invention since, a priori, an echography image provides less valuable
information than a MRI or scanner-type apparatus. Indeed, echography
usually provides a series of plane and independent cross-sectional images
of an organ instead of a volume image structured in series of parallel
image slices.
Reminders on Echography
Echography is an imaging process that has been progressing since 1970. A
transducer (piezoelectric crystal) emits ultra-sonic frequencies of
several megahertz that spread in the human body, but can be reflected when
they reach interfaces where the acoustic impedance of the medium presents
high variations (typically, a water/grease interface). The same transducer
can be used for a short period of time as an ultra-sonic frequency
emitter, and, for a generally longer period of time, as a receiver for the
ultrasonic frequencies reflected by the human body. It is then possible to
measure both the time interval between the emission and reception (which
allows, taking into account hypotheses on the speed of ultra-sonic
frequencies in the considered medium, the localization of the echography
source) and the echo intensity (which provides information on the nature
of the echographing point).
The simplest operation modality is to emit and recover ultra-sonic
frequencies in one direction only (mode A). Then, one obtains a signal
that is variable as a function of time if the body's tissues targeted by
echography are mobile. This operation mode was the first one to be used in
medicine, particularly in cardiovascular applications (where it made it
possible to assess the mobility of the hearth valves, for example).
It is also possible to emit and collect the ultrasonic frequencies in a
plane portion of the space (mode B). This can be achieved by juxtaposing
on a lineal rod a series of fixed mono-dimensional transducers, or by
rotating (mechanically or electronically), within a plane, about a fixed
point, a mono-dimensional transducer, Such an imaging modality has proven
very valuable for the study of "soft organs", more particularly in
gynecology, obstetrics and gastroenterology field.
Additionally, there exist many clinical situations in which echography
provides the most valuable information (more particularly in the
gastroenterology, gynecology-obstetrics and cardiology domains).
Echographies are also very useful for guiding surgical procedures (they
permit, for example, to control the introduction of puncture needles
through the human body).
Additionally, the echography system has over an X-ray scanner and MRI
apparatus the following advantages:
its cost is approximately 10 times lower,
the emitter can be fabricated in the form of a probe that is pressed on the
body, and has a light and easily transportable structure; on the contrary,
the X-ray scanner and MRI apparatuses are very cumbersome and occupy a
large volume in an examination room,
as compared to the X-ray scanner, echography, like MRI, is absolutely
harmless.
Above all, the main advantage of an echographic system to provide
morphologic images lies in its simplicity of use and low cost.
Additionally, usual preconceptions disregard the echography system for
orthopedics analyses, that is, for analyses of bones. Indeed, under common
conditions, ultra-sonic frequencies do not pass through the bones that
specularly reflect them. Therefore, it is impossible to study the inner
portion of bones by using ultrasonic frequencies. The study of a bone
surface is however possible, although it is rendered difficult because of
the specular character of the reflection of ultra-sonic frequencies. Since
the reflection is made practically only along the direction given by the
Cartesian law, each echography of a bone gives images including little
valuable information: only are seen portions of the outline of the bone
whose normal is parallel to the direction of the ultra-sonic frequencies.
However, for the application that is envisaged here, it is sufficient, as
will be described hereinafter, to collect a piece of information on a
point of the surface of a bone at each echography session to implement the
method according to the invention. Echography therefore also applies to
orthopedics applications.
More particularly, the invention provides a method for determining the
position of a patient's organ with respect to at least two imaging
devices. The second imaging device is a device providing an image of
points of the surface of an organ or of a skin region of the patient. This
method includes the steps consisting in determining the position of the
second device with respect to the coordinate system of the patient's
support; making at least one cliche with the first device, whereby a
second image corresponding to a cloud of points of the surface of the
organ or of the skin region is obtained; making at least one 3D image of
the organ and of its surface or of the surface of the skin region with the
first imaging device; and matching the second image with the first image.
According to an embodiment of the invention, the second imaging device is
an echography probe, capable of providing, at each view, the distances
within a viewing plane between the probe and the interface between the
organ and the adjacent medium. The imaging technique consists in moving
the probe and making a cliche for each of the plurality of predetermined
positions of the probe so as to obtain, for each cliche, the position with
respect to the probe of at least one point of the surface of the organ,
whereby a first image corresponding to a cloud of points of the organ
surface is obtained.
According to an embodiment of the invention, the step of determining the
position of the probe with respect to the coordinate system of the
patient's support includes the steps consisting in fixing the probe onto
an articulated arm of a coded robot, viewing three points of an echography
sighting mark, rotating the probe by 180.degree. about its viewing axis
and viewing again the three points of the sighting mark, and resuming the
two preceding steps for another position of the sighting mark.
According to an embodiment of the invention, the sighting mark is formed by
three parallel threads tightened between two planes, and by three
additional threads that are made of a material reflecting ultrasonic
frequencies, and arranged in a triangle abutting against the three
parallel threads. The set of threads is immersed in a medium capable of
transmitting ultrasonic frequencies, such as a water vessel.
The invention provides a method for determining the position of the
patient's organ in the on-site coordinate system in order to automatically
position a tool disposed within this coordinate system according to a
strategy, i.e. an orientation and a direction, that is defined by the
results of the analysis of the organ made in a pre-operation coordinate
system.
In this case, the second device is disposed on site, in an operation site
associated with a first coordinate system, the first imaging device is
disposed in a preoperation site associated with a second coordinate
system, and the first imaging device is of a type adapted to provide a 3D
image of the surface of the organ, and, if required, also of the skin
surface, whereby the image matching step localizes the first coordinate
system with respect to the second one.
A embodiment of the invention provides the steps consisting in determining
an action axis of a surgical tool with relation to the pre-operation image
taken in the first coordinate system, identifying this action axis in the
first coordinate system, and positioning a tool according to this action
axis in the second coordinate system. When the "surgical" tool is an
analysis system mounted on a support placed in the operation site but
independent of the support of the first device, the identification of the
tool's action axis is made by viewing a same sighting mark by the second
device and by the tool. When the second device is an echography probe and
the tool is a radiotherapy apparatus having its sphere of action
materialized by laser beams, the sighting mark that is common to the
echography probe and to this tool is an echography sighting mark to which
is added a cubic reflector positioned in a determined manner with respect
to the latter.
The invention also provides to use this method to position a morphologic
image with respect to a functional image of an organ, the functional image
being made, for example, by a gamma-scintigraphy camera or a PET
apparatus.
In this case, the second device is positioned in an operation site, and a
third imaging device is used, such as a gamma-scintigraphy camera or PET,
MEG or synchrotron-radiation apparatus, placed in the same operation site.
The relative positions of the second device and third devices are defined
by viewing a same sighting mark by the second and third devices. The image
matching is made from this initial determination of the relative positions
of the first and third devices.
According to an embodiment of the invention, in which the second device is
an echography probe, and the third device is a gamma-scintigraphy or MEG
apparatus, the sighting mark is formed by hollow catheters filled with a
radioactive product, the sighting mark including four tubes arranged in a
noncollinear manner between two parallel plates.
The foregoing and other objects, features, aspects and advantages of the
invention will become apparent from the following detailed description of
the present invention when taken in conjunction with the accompanying
drawings.
BRIEF DESCRIPTION OF DRAWINGS
FIG. 1 illustrates an exemplary sighting mark used with an echography
system according to the invention; and
FIG. 2 schematically represents an exemplary dual
echography/gamma-scintigraphy sighting mark used according to the
invention.
DETAILED DESCRIPTION
As reminded above, for several medical, diagnostic or therapeutic
procedures, it is necessary to match (that is, to make correspond in a
determined manner) a first coordinate system in which a previous,
pre-operation, examination has permitted to study a portion of the anatomy
and to determine a surgical procedure, and a second, on-site, coordinate
system in which the surgical procedure is carried out. The prior art
methods mainly consist in fixing fiducial marks visible both during the
surgical procedure and during the pre-operation examination. As indicated
above, such methods are often inaccurate and difficult to use.
The pre-operation examination must allow the identification of anatomic
portions (a vertebra, a prostate, the brain . . . ) whose shape or the
shape of the skin envelope (the skin of the skull for the brain) will
serve as a reference for matching the coordinate systems. Thus, one
considers the case where the coordinate systems are matched ("merged") as
a function of the organ itself or of its skin envelope. Various image
processing techniques can be used to implement these operations by using a
so-called 3D segmentation step (refer to F. Leitner, I. Marque. S.
Lavallee, P. Cinquin, Dynamic Segmentation: "Finding the Edges with Spline
Snakes", Proc. Int. Conf. on Curves and Surfaces, Chamonix, Academic
Press, pp. 279-284, (1991).
Using an echography probe
In an embodiment, the present invention aims at providing a method in which
the on-site image results from a cloud of points obtained by echography
examination of the region of interest, which permits visualizing objects
that have been previously segmented.
The difficulty lies in the conception of a protocol that enables to
associate the coordinates of the points that were observed in echography
with the on-site coordinate system. To achieve this purpose, one must be
capable of localizing the position of the echography probe in the on-site
(surgical) coordinate system.
According to a first implementation of the invention, it is possible to
provide on the probe itself landmarks detectable by an adequate sensor
(for example, photoluminescent diodes, ultra-sonic frequency emitters)
that is rigidly fixed with respect to the on-site coordinate system.
According to another preferred embodiment of the invention, the probe is
rigidly fixed to the end of an articulated arm of a robot. Then, one
determines both the position of the probe with respect to the articulated
arm and the position of the coordinate system of the articulated arm with
respect to the on-site coordinate system.
a) Determining the relative position of the probe with respect to the
articulated arm
To achieve this determination, the invention provides for determining the
position of the probe with respect to a calibration sighting mark that
permits to see in echography landmarks having a fixed spatial
distribution. These landmarks are scanned by the echography device and
their actual spatial position is compared with the position provided by
the coordinate transformer of the robot's arm for a theoretical position
of the probe. Then, by using a non-linear least square technique, it is
possible to identify rotation and translation parameters that characterize
the transition from the coordinate system of the probe to the one of the
arm.
Since the probe is rigidly fixed to the arm's end, the transformation that
associates the coordinate system of the probe with the coordinate system
of the articulated arm must be found. To achieve this purpose, three
reference points of a sighting mark, that have to be studied in at least
two arbitrary positions, are studied.
One embodiment of a sighting mark according to the invention is illustrated
in FIG. 1. The sighting mark includes, in a medium capable of transmitting
ultrasonic frequencies, for example a water vessel 10, three threads 1, 2
and 3 tightened between two planes. Three additional threads, 4, 5, and 6
connect each couple of the three threads 1, 2 and 3, and form a triangular
pattern. The triangle can be fabricated by a thin thread made of a
material sensitive to echography system, such as a Nylon thread. Then, by
means of the articulated arm manipulator (not shown), a probe 11 is
arranged so that its plane beam 12 is coplanar with the plane of the three
threads 4, 5 and 6. When the probe is within the triangle's plane, this
triangle is then perfectly visualized and its apex is deducted by
calculating the intersection of its edges. From this position of probe 11,
the probe is rotated by 180.degree. about axis Z comprised within the
plane of the echography image, thus allowing to identify the rotation
parameters.
The calibration sighting mark is moved to another arbitrary position so as
to resume the visualization of the points of the sighting mark according
to two positions of the articulated arm supporting the echography probe,
the two positions being rotated one with respect to the other by
180.degree.. Then, all the necessary data are available to implement a
conventional calibration method, such as described, for example, by Y. C.
Shiu et al., "Finding the Mounting Position by Solving a Homogeneous
Transform Equation of Form AX=XB", CH24133/87/0000/1666, 1987, IEEE.
b) Determining the position of the fiducial mark of the articulated arm and
of the probe with respect to the coordinate system of a surgical tool.
The surgical tool, for example a guide for introducing a needle, may have
been worked simultaneously with the articulated arm carrying the probe. In
that case, there is no particular positioning problem.
However, if the surgical tool was designed independently of the articulated
arm carrying the probe, their coordinate sys | | |
