U.S. patent application number 15/076232 was filed with the patent office on 2016-07-14 for reducing incremental measurement sensor error.
The applicant listed for this patent is Hansen Medical, Inc.. Invention is credited to June Park, Neal A. Tanner, Sean P. Walker, Serena H. Wong.
Application Number | 20160202053 15/076232 |
Document ID | / |
Family ID | 51523443 |
Filed Date | 2016-07-14 |
United States Patent
Application |
20160202053 |
Kind Code |
A1 |
Walker; Sean P. ; et
al. |
July 14, 2016 |
REDUCING INCREMENTAL MEASUREMENT SENSOR ERROR
Abstract
For position sensors, e.g., a fiber-based system, that build a
shape of an elongated member, such as a catheter, using a sequence
of small orientation measurements, a small error in orientation at
the proximal end of the sensor will cause large error in position
at distal points on the fiber. Exemplary methods and systems are
disclosed, which may provide full or partial registration along the
length of the sensor to reduce the influence of the measurement
error. Additional examples are directed to applying selective
filtering at a proximal end of the elongated member to provide a
more stable base for distal measurements and thereby reducing the
influence of measurement errors.
Inventors: |
Walker; Sean P.; (Fremont,
CA) ; Wong; Serena H.; (Mountain View, CA) ;
Park; June; (Palo Alto, CA) ; Tanner; Neal A.;
(Burnet, TX) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Hansen Medical, Inc. |
Mountain View |
CA |
US |
|
|
Family ID: |
51523443 |
Appl. No.: |
15/076232 |
Filed: |
March 21, 2016 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14712587 |
May 14, 2015 |
9289578 |
|
|
15076232 |
|
|
|
|
14208514 |
Mar 13, 2014 |
9057600 |
|
|
14712587 |
|
|
|
|
61779742 |
Mar 13, 2013 |
|
|
|
Current U.S.
Class: |
250/227.14 |
Current CPC
Class: |
G01B 11/24 20130101;
A61B 5/0066 20130101; A61M 25/0108 20130101; A61B 6/12 20130101;
A61B 6/586 20130101; A61B 2090/3966 20160201; A61M 2205/702
20130101; A61B 6/485 20130101; A61B 6/032 20130101; G01B 11/14
20130101; A61B 90/39 20160201; A61B 8/12 20130101; A61M 2025/0166
20130101; G01B 21/045 20130101 |
International
Class: |
G01B 21/04 20060101
G01B021/04; G01B 11/24 20060101 G01B011/24; A61B 6/00 20060101
A61B006/00; A61B 6/12 20060101 A61B006/12; A61B 6/03 20060101
A61B006/03 |
Claims
1. (canceled)
2. A method for registering a tool to a pre-operative
three-dimensional model with the use of an imaging sensor and a
localization sensor, the method comprising: obtaining imaging data
from an imaging sensor coupled to a tool moving within a blood
vessel; obtaining localization data from the localization sensor;
generating an image of the blood vessel from the imaging data and
localization data; correlating the generated image to a
pre-operative three-dimensional model of the blood vessel to
register the tool to the pre-operative three-dimensional model.
3. The method of claim 2, wherein the generated image comprises a
sequence of images or a three-dimensional reconstruction.
4. The method of claim 2, wherein the localization data comprises
position or shape data.
5. The method of claim 2, wherein the imaging sensor is an
intravascular ultrasound probe.
6. The method of claim 5, wherein the image generated by the
intravascular ultrasound probe is a two-dimensional cross-sectional
view of the blood vessel.
7. The method of claim 2, wherein the imaging sensor is an optical
coherence tomography sensor.
8. The method of claim 7, wherein the image generated by the
optical coherence tomography sensor is a local three-dimensional
view of the blood vessel.
9. The method of claim 2, wherein generating the image of the blood
vessel from the imaging data and localization data and correlating
the generated image to a pre-operative three-dimensional model are
performed by implementing a single recursive filtering
algorithm.
10. The method of claim 9, wherein the single recursive filtering
algorithm utilizes a Bayesian filter to estimate a position of the
tool relative to the pre-operative three-dimensional model.
11. The method of claim 10, wherein the Bayesian filter comprises
an Extended Kalman Filter, an Unscented Kalman Filter, or a
particle filter.
12. The method of claim 9, wherein implementing the single
recursive filtering algorithm comprises predicting a motion of the
tool in the pre-operative three-dimensional model, and performing
an update of the prediction at each time step based on new sensor
measurements.
13. The method of claim 12, wherein performing the update of the
prediction at each time step is performed using a kinematic
model.
14. The method of claim 13, wherein performing the update of the
prediction at each time step comprises applying a correction to a
Bayesian filter state by comparing a predicted vessel wall position
to a sensed position of the blood vessel.
15. The method of claim 2, further comprising displaying the
position and shape of the tool within the pre-operative
three-dimensional model without using intra-operative
fluoroscopy.
16. The method of claim 2, further comprising sensing an
environment around the tool using the imaging sensor, and
positioning and displaying sensed environmental features relative
to the pre-operative three-dimensional model.
17. The method of claim 16, wherein displaying the sensed
environmental features relative to the pre-operative
three-dimensional model comprises providing a real-time view of an
output of the imaging sensor alongside a view of the tool
positioned in the pre-operative three-dimensional model.
18. The method of claim 17, wherein displaying the sensed
environmental features relative to the pre-operative
three-dimensional model comprises superimposing the sensed
environmental features on the pre-operative three-dimensional
model.
19. The method of claim 2, further comprising processing the image
to obtain an estimate of a curve or surface representing a wall of
the blood vessel.
Description
CROSS REFERENCE TO RELATED APPLICATIONS
[0001] This application is a continuation of U.S. patent
application Ser. No. 14/712,587, filed May 14, 2015, which is a
continuation of U.S. patent application Ser. No. 14/208,514, issued
as U.S. Pat. No. 9,057,600, filed Mar. 13, 2014, which claims
benefit of U.S. Provisional Patent Application No. 61/779,742,
filed Mar. 13, 2013. The contents of all of the above-referenced
patent applications are hereby incorporated by reference in their
entirety.
BACKGROUND
[0002] Currently known minimally invasive procedures for diagnosis
and treatment of medical conditions use shapeable instruments, such
as steerable devices, flexible catheters or more rigid arms or
shafts, to approach and address various tissue structures within
the body. For various reasons, it is highly valuable to be able to
determine the 3-dimensional spatial position of portions of such
shapeable instruments relative to other structures, such as the
operating table, other instruments, or pertinent anatomical tissue
structures. Such information can be used for a variety of reasons,
including, but not limited to: improve device control; to improve
mapping of the region; to adapt control system parameters (whether
kinematic and/or solid mechanic parameters); to estimate, plan
and/or control reaction forces of the device upon the anatomy;
and/or to even monitor the system characteristics for determination
of mechanical problems. Alternatively, or in combination, shape
information can be useful to simply visualize the tool with respect
to the anatomy or other regions, whether real or virtual.
[0003] In many conventional systems, the catheter (or other
shapeable instrument) is controlled in an open-loop manner, as
described in U.S. patent Ser. No. 12/822,876, issued as U.S. Pat.
No. 8,460,236, the contents of which are incorporated by reference
in its entirety. However, at times the assumed motion of the
catheter does not match the actual motion of the catheter. One such
reason for this issue is the presence of unanticipated or
un-modeled constraints imposed by the patient's anatomy. Another
reason for this may be that the parameters of the tool do not meet
the ideal/anticipated parameters because of manufacturing
tolerances or changes in the mechanical properties of the tool from
the environment and aging.
[0004] Thus to perform certain desired applications, such as, for
example, instinctive driving, shape feedback, and driving in a
fluoroscopy view or a model, there exists a need for tool sensors
to be properly registered to the patient in real time. Moreover,
there remains a need to apply the information gained by spatial
information or shape and applying this information to produce
improved device control or improved modeling when directing a
robotic or similar device. There also remains a need to apply such
controls to medical procedures and equipment.
[0005] Localization sensors such as fiber optic shape sensors may
include Incremental Measurement Sensors (IMSs). An IMS measures a
shape or path of an elongate member by combining a sequence of
serial orientation and distance measurements. For instance, FOSSL
generates a shape by measuring types of strain at discrete points
in the fiber; this strain is then translated to the incremental
change in roll and bend, which is incremented along all steps to
obtain the position and orientation at a given location. As a
result, each position and orientation at a point is dependent on
the position and orientation of all proceeding points. In contrast,
an electromagnetic coil sensor measures position at points along
the elongate member independent of any other measurements.
[0006] One drawback of IMSs is that a measurement noise (error) at
any location along the path may propagate to all measurements
distal to that measurement. While these errors are implicit in the
nature of the sensor, orientation errors at a proximal portion of
the IMS may result in a large position error at the distal end of
the elongate member. In applications that include accurate distal
position measurements, this can cause the measured tip position to
vary greatly between successive measurements due to noise at a
single point in the proximal portion. One way of thinking about the
issue is to consider the IMS length as a lever arm--small rotations
at one end cause large changes in the position at the other end.
The longer the lever arm, the more pronounced the conversion from
proximal orientation error to distal position error. It should be
noted that an orientation error at the proximal end will not tend
to cause a large orientation error at the distal end because
orientation errors themselves accumulate (sum) over the length of
the sensor.
[0007] Thus, for Incremental Measurement Sensors that build a shape
using a sequence of small orientation measurements, a small error
in orientation at the proximal end of the sensor will cause a large
error in position at distal points on the fiber. Accordingly, there
is a need for an improved method of using IMSs that reduces
measurement errors.
SUMMARY
[0008] Exemplary systems and methods are disclosed for reducing
measurement error, e.g., relating to position measurements of an
elongated member, e.g., along a distal portion of the elongated
member. An exemplary method includes providing an incremental
sensor measurement at a distal position on an elongated member, and
applying registration data at one or more proximal locations along
the elongated member. This exemplary method may further include
determining a position of the incremental measurement sensor based
at least upon the registration data from the one or more proximal
locations.
[0009] In another exemplary method, either alternatively or in
addition to the above-described registration data, a proximal
signal of the elongated member may be selectively filtered, e.g.,
in comparison to a distal portion of the elongated member. In such
examples, a distal position of the incremental measurement sensor
may be determined based at least upon the filtered proximal signal,
thereby reducing a fluctuation of the determined distal position of
incremental measurement sensor.
[0010] An exemplary measurement system may include an incremental
sensor measurement positioned at a distal position on an elongated
member, and a processor. In some exemplary approaches, the
processor may be configured to apply registration data at one or
more proximal locations along the elongated member, and to
determine a position of the incremental measurement sensor based at
least upon the registration data from the one or more proximal
locations. In other examples, either alternatively or in addition
to relying upon registration data, the processor may be configured
to selectively filter a proximal signal of the elongated member,
and determine a position of the incremental measurement sensor
based at least upon the filtered proximal signal.
BRIEF DESCRIPTION
[0011] While the claims are not limited to a specific illustration,
an appreciation of the various aspects is best gained through a
discussion of various examples thereof. Referring now to the
drawings, exemplary illustrations are shown in detail. Although the
drawings represent the illustrations, the drawings are not
necessarily to scale and certain features may be exaggerated to
better illustrate and explain an innovative aspect of an example.
Further, the exemplary illustrations described herein are not
intended to be exhaustive or otherwise limiting or restricted to
the precise form and configuration shown in the drawings and
disclosed in the following detailed description. Exemplary
illustrations are described in detail by referring to the drawings
as follows:
[0012] FIG. 1A illustrates a variation of a localization system in
a typical operation room set up.
[0013] FIG. 1B illustrates a 3D Model frame.
[0014] FIG. 2 illustrates an exemplary robotic surgical system.
[0015] FIG. 3 is a schematic representation of a first registration
technique of correlating a sensor reference frame to selective
reference frames.
[0016] FIG. 4 is a flow chart that illustrates a method of
transforming a reference frame for a sensor of a surgical tool into
a target reference frame.
[0017] FIG. 5 is a flow chart that illustrates a method of
transforming a reference frame associated with a tool into a target
reference frame.
[0018] FIG. 6 is a flow chart that illustrates a method of
transforming a reference frame associated with a tool into a target
reference frame utilizing medical appliances.
[0019] FIG. 7 is a flow chart that illustrates a method of using a
sensor to transform a reference frame associated with a tool into a
target reference frame.
[0020] FIG. 8 is a schematic illustration of a method of using an
intravascular imaging sensor coupled with a shape sensor to
transform a reference frame associated with a tool into a target
reference frame.
[0021] FIG. 9 is a schematic illustration of an exemplary elongate
member, e.g., a fiber, having a proximal end and a distal end.
[0022] FIG. 10 is a schematic illustration of another exemplary
elongate member having a distal end inserted into a patient, with a
proximal end remaining outside the patient.
[0023] FIG. 11 is a process flow diagram for an exemplary process
of reducing measurement error associated with the position of an
elongate member.
DETAILED DESCRIPTION OF THE INVENTION
[0024] Exemplary approaches described herein provide full or
partial registration along the length of a sensor to reduce the
influence of measurement error. As will be described further below,
an exemplary registration process generally relates a reference
frame of a sensor to another reference frame of interest.
Alternatively or in addition to examples employing full or partial
registration along the length of a sensor, applying heavier
filtering at the proximal end of the sensor can provide a more
stable base for distal measurements reducing the influence of
measurement errors. While these exemplary approaches are discussed
in the context of a Fiber Optic Shape Sensing and Localization
(FOSSL) system, other visualization methodologies may be employed.
Fiber optic shape sensing is a technology that can sense the shape
of a flexible body such as a catheter during a surgical procedure
to permit visualization of the catheter in the patient's
anatomy.
[0025] One exemplary methodology provides a solution to this
problem by using point registration along the length of the fiber
to reduce the effect of the lever arm problem. As noted above,
these techniques apply to all sensors that use incremental
measurements, but nevertheless exemplary approaches below are
described generally in the context of FOSSL fiber technology.
[0026] Another exemplary approach includes applying registration
data along the fiber as well as at the proximal attachment to
reduce error. Additionally, in another example stronger filtering
may be utilized on the proximal orientation signal to reduce
noticeable fluctuation of the position distally.
[0027] In still another exemplary illustration, orientation error
at the proximal end of an IMS is decoupled from position error at
the distal end. Exemplary methods of decoupling this error include
providing some notion of a registration closer to the distal tip
and reduce the effect of orientation error.
[0028] Various localization systems and methods for tracking an
elongate instrument or tool, e.g., a robotically controlled
elongate instrument, in real time, in a clinical or other
environment, are described herein. The term "localization" is used
in the art in reference to systems for determining and/or
monitoring the position of objects, such as medical instruments or
tools in a reference coordinate system. Various instruments are
contemplated for use in the various systems described herein. In
one exemplary arrangement, elongate instruments are contemplated,
such as, e.g., a catheter or vascular catheter. The various methods
and systems may include integrating or registering a localization
system or a localization sensor coupled to a surgical tool, with an
image. A fiber optic tracking or localization system is just one,
non-limiting example of a system that allows for the tracking of a
location, position and/or orientation of a localization sensor
placed. Various other localization sensors may be utilized, e.g.,
electromagnetic sensors, and other sensors for detecting or
controlling the movement of medical equipment. When the
localization sensor is integrated into an image, it enhances the
capabilities of an instrument control or tracking system by
allowing a user or doctor to easily navigate the instrument through
the complex anatomy without exposing the patient to excessive
radiation over a prolonged period of time.
[0029] The localization data or tracking information of a
localization sensor may be registered to the desired image or model
to allow for navigation of an elongate instrument through the image
or model to accurately represent movement of the elongate
instrument within a patient. As noted above, registration is a
process that generally requires relating a reference frame of a
sensor to another reference frame of interest. If the positions,
orientations or shapes of two or more objects are known in the same
reference frame, then the actual positions, orientations or shapes
of each object relative to each other may be ascertained. Thus,
with this information, one can drive or manipulate one of the
objects relative to the other objects.
[0030] In most interventional procedures, the reference frame of
interest is the visualization frame. The reference frame is the
frame that the doctor is viewing, such as a patient or a live 2D/3D
image such fluoroscopy, ultrasound or others. Thus, the goal of
registration is to determine the relationship of a frame of a
sensor integrated into a tool or element in the surgical suite
within the frame of reference of the patient, as represented in a
2D/3D image.
[0031] When the tool is registered to a 3D model, the user can
drive and manipulate the tool in the 3D model. This technique
provides an advantage in that there is no longer a need for live
fluoroscopy and radiation during a procedure. The tool is localized
to the 3D model and the position, shape and orientation of the tool
is visible to the user. Since the tool position, shape and
orientation is updated in real time by a localization sensor, an
image of the tool in the virtual representation of the 3D model
will be updated as it is being advanced into the patient. The
sensor is localized to the reference frame of the 3D model;
therefore the orientation of a tip of the tool is known relative to
the 3D model. This enables driving of the tool (such as a catheter)
within 3 dimensional views of the anatomy and hence improves
visualization and control during a surgical procedure.
[0032] As discussed above, exemplary sensors may include
incremental measurement sensors, where the position and orientation
of a particular point is calculated and dependent on the previously
calculated orientations and positions of proximal points or
(spacing of consecutive points). Thus, the localization sensor
operating in any medical system needs to be registered with a
coordinate system, frame or image that is useful to an operator,
such as the pre-operative 3D model or a fluoroscopic image. For
incremental measurement sensors, such registration is challenging
because the coordinate system or frame of the sensor is not always
easily related to the coordinate system of interest (i.e., the
pre-operative 3D model).
[0033] Moreover, the relationship between the sensor and the
coordinate system of the interest may change over time during a
procedure. For example, in one exemplary robotic system, a fiber
optic sensor may have its reference frame based physically in a
splayer (base) for a catheter. Thus, as the splayer is robotically
driven during a surgical procedure, the position and orientation of
the base of the fiber will change with respect to other reference
frames.
[0034] In addition to changing positions of reference frames, the
registration process often requires information about the imaging
system providing the image, such as its physical dimensions and/or
the details about the imaging techniques used to acquire a
particular 3D model or other image. Due to the variability in
equipment used in a clinical environment, in certain situations
there may be no guarantee that such information will be available
or easily obtainable to an outside party.
[0035] As such, various techniques to estimate system parameters
and various registration techniques may help facilitate the
clinical use of localization technology.
[0036] In certain variations, a method for tracking a robotically
controlled elongate instrument in real time may include displaying
an image of a patient's anatomy. A localization sensor may then be
coupled to the robotically controlled instrument. The localization
sensor may provide localization data of the sensor and/or
instrument. Moreover, different sensors may be registered to
specific tools, thereby enabling tool differentiation. The
localization data from the localization sensor may be registered to
the image. Registering may include transforming localization data
generated by the localization sensor to the coordinate system or
frame of the image such that localization data of the elongate
instrument, to which the localization sensor is coupled, is
overlaid on the image. The coordinate system of the localization
sensor may be transformed or translated to the coordinate system of
the image through one or more transformations, and optionally
through other coordinate systems, to register the localization data
to the image. As a result, a continuously or substantially
continuously updated location of at least a portion of the elongate
instrument is provided in the image of the anatomy of a patient,
which allows for or facilitates robotic navigation or control of
the elongate instrument through the anatomy e.g., through the
vasculature of a patient.
[0037] The location, position and/or orientation of the
localization sensor may be continuously tracked to allow for
accurate manipulation of the elongate instrument in or through the
anatomy of a patient. Various types of images may be utilized in
the methods and systems described herein. For example, an image may
be generated by CT or 2D or 3D fluoroscopy. An image may include a
3D or 2D anatomical model or a 2D or 3D fluoroscopic image or other
types of images useful for visualizing an anatomy of a patient to
perform various medical procedures.
[0038] When using a fluoroscopy image, an image intensifier may be
utilized. Localization data from the localization sensor may be
registered to a fluoroscopy coordinate system of a fluoroscopy
image coupled to the image intensifier. In order to register the
localization data from the localization sensor to the fluoroscopy
image, various parameters may be ascertained or known. For example,
such parameters may include: a distance from an X-ray source to the
image intensifier, a distance from the source to a bed, a size of
the image intensifier, and/or the axis of rotation of a C-arm of
the fluoroscopy system.
[0039] In certain variations, localization data can be registered
to a 3D anatomical model or a fluoroscopy image. The techniques
used to perform the registration vary depending on the target.
Where localization data is registered to a fluoroscopy image, the
2D nature of the fluoroscopy images may require that multiple
images be taken at different angles before the registration process
is complete.
[0040] FIG. 1A is a schematic of a typical operation room set up
for a robotic surgical system. More specifically, a typical robotic
surgical system 10 includes a table 12 upon which a patient 14 will
be placed, a fluoroscopy system 16, and a surgical tool, such as a
catheter 18 (best seen in FIG. 2). Attached to the table 12 is a
setup joint arm 20 to which a remote catheter manipulator (RCM) 22
is operatively connected. A splayer 24 may be mounted to the RCM
22. A surgical tool, such as a catheter, is operatively connected
to the splayer 24. A fiber sensor 26 may be operatively connected
to the surgical tool. The fluoroscopy system 16 includes a C-arm
28. A fluoroscopy panel 30 is mounted to the C-arm 28. The C-arm is
selectively moveable during the procedure to permit various images
of the patient to be taken by the fluoroscopy panel 30.
[0041] Additional portions of the robotic surgical system 10 may be
further seen in FIG. 2. More specifically, robotic surgical system
10 may further comprise an operator control station 31, which may
be remotely positioned with respect to table 12. A communication
link 32 transfers signals between the operator control station 31
and the RCM 22. The operator control station 31 includes a control
console 34, a computer 36, a computer interface, such as a mouse, a
visual display system 38 and a master input device 40. The master
input device 40 may include, but is not limited to, a multi-degree
of freedom device having multiple joints and associated
encoders.
[0042] Each element of the robotic surgical system 10 positioned
within the operating suite may define a separate reference frame to
which sensors may be localized. More specifically, separate
reference frames may be defined for each of elements of the robotic
surgical system 10. Such reference frames may include the
following: a table reference frame TRF for the table 12, a setup
joint frame SJF for the setup joint 20, an RCM reference frame RRF
for the remote catheter manipulator (RCM) 22, a splayer reference
frame SRF, a fluoroscopy reference frame FF. Separate reference
frames may also be defined for a patient--patient reference frame
PRR, a reference frame FRF for a sensor disposed within a surgical
tool, and a pre-operative 3D frame AMF (best seen in FIG. 1B).
[0043] To relate a coordinate frame of a fiber optic sensor of a
tool to either a fluoroscopy frame FF, or a pre-operative 3D frame
AMF, a variety of registration techniques may be employed.
Generally, the techniques proposed herein fall into several
categories. A first category involves using image processing or
vision techniques to relate a reference frame RFR of a fiber sensor
directly to an image or 3D model. This technique may be
accomplished manually by a user or done automatically using image
processing techniques. Another category to coordinate the reference
frame FRF of a fiber optic sensor involves using knowledge about
hardware, and potentially other sensors and or position of the
fiber. Further discussion of these techniques is set forth
below.
Registration to Fluoroscopy Coordinate Frame
[0044] Referring to the systems illustrated in FIGS. 1-3, the first
category of registration techniques will now be described. The
first category relates the coordinate system of the sensor
reference frame FRF to a fluoroscopy reference frame FF directly.
This technique utilizes fluoroscopy images taken during the
surgical procedure by the fluoroscopy system 30, in real-time.
[0045] More specifically, an exemplary registration process 200 is
illustrated in the flow chart of FIG. 4. The process 200 may begin
by inserting a tool into a patient at block 202. As described
above, in one exemplary configuration, the tool is a catheter 18,
which may be inserted by an RCM 22. Next, at block 204 an
intra-operative image is taken of the tool 18.
[0046] In one exemplary arrangement, the intra-operative image is a
fluoroscopy image taken by fluoroscopy system 30. Next, distinctive
elements of the tool may be identified in the fluoroscopy image at
block 206. In one exemplary configuration, the block 206 may be
accomplished by instructing the user to select certain marked
points of a catheter 18 in the fluoroscopy image at the work
station 31. Examples of marked points include, but are not limited
to, physical features of the catheter 18 such as the tip of the
catheter 18, certain shapes and an articulation band. In other
exemplary configurations, fluoroscopy markers may be disposed on
the catheter.
[0047] Once the selected points are identified in the fluoroscopy
image, in the next step 208, coordinates of the selected points of
the catheter 18 may be compared to corresponding measured points of
elements of the catheter. In one exemplary configuration, measured
points from a tool sensor operatively connected to the tool 18 may
be used. More specifically, in one exemplary configuration, the
tool sensor is a fiber optic sensor. Information about the fiber
optic sensor will be known in relation to the features on the
catheter from an in-factory calibration. This comparison can be
used to determine a transformation matrix that can be used to
transform a reference frame FRF for a sensor disposed within the
surgical tool into the fluoroscopy reference frame FF. This
transformation then localizes the tool relative to the
intra-operative fluoroscopy image.
[0048] Once the fiber sensor of the tool has been registered or
localized to the fluoroscopy image, the tool operator can now move
or drive the tool to various, desired points visualized in the
fluoroscopy image. Moreover, the computer 36 may be configured to
track the marked points over time, such that an appropriate
transformation may be updated.
[0049] In one exemplary configuration, the identifiable markers
need not be on the portion of the tool that is inserted into the
patient. For example, markers may be embedded on a splayer 24,
which may allow for larger and more complex markers to provide
enhanced registration capabilities.
[0050] As described above, in addition to utilizing fluoroscopy
marked points, it is also contemplated that distinct shapes that
may be visible under fluoroscopy may also be used. However, this
technique will require some image segmentation (to identify and
separate out targeted shapes).
[0051] With respect to the proposed technique of localizing a
sensor reference frame FRF to the fluoroscopy reference frame FF,
the localization sensor could serve to reduce the use of
fluoroscopy during a procedure. More specifically, the use of
fluoroscopy would be reduced since fluoroscopy would only be
required when re-registration is needed during the procedure due to
the loss of accuracy in the data obtained from the sensor.
[0052] In certain arrangements, it may be desirable to further
register the tool to a 3D model reference frame AMF, as illustrated
in FIG. 3. Registration to the 3D Model is discussed more fully
below.
Registration Through Successive Physical Components
[0053] Another exemplary technique proposed to register a tool 18
to a desired reference frame involves the use of physical
components of the medical system 10 and multiplying successive
transformations. This proposed technique 300 is illustrated
schematically in FIG. 5 and involves finding a transformation path
from a tool reference frame such as a fiber sensor, splayer 24, or
catheter 18, to the table 12, as in most surgical suite setups, the
table location is generally known with respect to the fluoroscopy
system 30. More specifically, registration technique 300 involves
determining a tool reference frame at block 302 (where the tool
reference frame may be defined as the sensor reference frame FRF,
splayer reference frame SRF or a catheter reference frame) and
correlating the tool reference frame to a table reference frame TRF
at block 304, thereby registering the tool 18 to the table 12.
Registering the tool 18 to the table 12 will serve to provide
necessary information to permit registration to an additional
target frame, such as a fluoroscopy reference frame FF, for
example. Because the table 12 location is typically known with
respect to a fluoroscopy system 30, a comparison of set reference
points of the table 12 with corresponding reference points in a
fluoroscopy image may be used to determine a transformation matrix
to transform the table reference frame TRF into the fluoroscopy
reference frame FF. The tool 18 is registered to the table
reference frame TRF, and thus combining all three transformations
localizes the tool relative to the intra-operative fluoroscopy
image.
[0054] However, it is understood that the present disclosure does
not require that the tool 18 be registered to the table 12. Indeed,
it is expressly contemplated that registration of the tool 18 to
other physical components within the surgical suite may also be
utilized. This proposed technique requires the use of other sensors
in addition to, or alternative to a fiber sensor, however.
Exemplary configurations of registration through physical surgical
suite components are discussed in further detail below.
[0055] One exemplary method of performing registration through
successive physical components is illustrated in the flow chart in
FIG. 6. In this technique, the registration process 400 begins at
block 402, with determining the location of the setup joint 20 with
respect to the table 12. Encoders on the RCM 22 and setup joint 20,
with kinematic models may be used to determine the location of the
setup joint 20 with respect to the table 12. More specifically, the
encoders assist with determining the location of the RCM 22 with
respect to the table 12. With the location value of the position
that the setup joint 20 is fixed to the table 12, the location of
the splayer carriage 24 carried by the RCM 22 with respect to the
table 12 can be determined; i.e., the setup joint reference frame
SJF is localized with the RCM reference frame RRF. Because
information about the catheter will be known in relation to the
splayer carriage 24 from an in-factory calibration, at block 404 of
the registration process 400, an evaluation of the splayer carriage
24 information with respect to the RCM can be used to determine a
transformation matrix that can be used to transform the splayer
carriage reference frame SRF to the table reference frame TRF. As
described above, because the table 12 location is known with
respect to the fluoroscopy system 30, at block 406 another
transformation may be done from the table reference frame TRF to
the fluoroscopy reference frame FF. This final transformation,
i.e., from the table reference frame TRF to the fluoroscopy
reference frame FF, then localizes the tool relative to the
intra-operative fluoroscopy image.
[0056] In another exemplary method of performing registration
through successive physical components, inertial sensors on the RCM
22, coupled with the information about the initial position of the
RCM 22 on the table 12, may be used to assist in localizing the
catheter splayer reference frame SRF to the table reference frame
TRF. More specifically, once the RCM 22 is localized to the table
reference frame TRF, the catheter splayer reference frame SRF may
be localized to the table reference frame TRF, as the position of
the catheter splayer 24 with respect to the RCM 22 will be known
from in-factory calibration.
[0057] Yet another exemplary method 500 of performing registration
through physical components is illustrated in FIG. 7. The method
500 uses a second fiber optic sensor. In a first step 502, one end
of the fiber optic sensor is fixed to the table 12. Next, in step
504, the other end of the sensor is fixed to the splayer 24 in a
known orientation/position. In this technique, a position and
orientation transformation between the tip and base of the fiber
sensor may be determined, thereby localizing the catheter splayer
reference frame SRF to the table reference frame TRF in step 506.
However, it is understood that the initial position of the fixed
point at the table must be known. Once the catheter splayer
reference frame SRF is localized to the table reference frame TRF,
because the table 12 location is known with respect to the
fluoroscopy system 30, in step 508 another transformation may be
done from the table reference frame TRF to the fluoroscopy
reference frame FF. This final transformation, i.e., from the table
reference frame TRF to the fluoroscopy reference frame FF, then
localizes the tool relative to the intra-operative fluoroscopy
image.
[0058] A further exemplary method of performing registration of a
surgical tool to a physical component includes using
electromagnetic sensors to track the location of the splayer 24
with respect to an electromagnetic sensor at a known location on
the table 12. In using this technique, because the tool location is
calibrated to the splayer 24 in the factory, once the splayer 24 is
localized to the table reference frame TRF, the tool may be
localized to the fluoroscopy reference frame FF as the table 12 is
known with respect to the fluoroscopy system 30.
[0059] In yet another exemplary method, instead of electromagnetic
sensors, overhead cameras or other visualization techniques may be
employed to track distinct features on the splayer 24 and the table
12 to determine the respective orientation and position with regard
to each other.
[0060] A further technique may use the range sensors (such as,
e.g., IR or ultrasound) to find the distance to several distinct
and predetermined points on the table 12 and the splayer 24. Once
the splayer 24 is localized to the table reference frame TRF, the
tool may be localized to the fluoroscopy reference frame FF as the
table 12 is known with respect to the fluoroscopy system 30.
[0061] All of the above techniques serve to register the tool to a
physical component within the surgical suite, such as, for example,
the table 12. Some of the above techniques require the RCM 22 and
setup joint 20 to be registered to the table 12. That
pre-registration step may be achieved by using some known feature
on the table 12 that the setup joint 20 may reference.
Additionally, the pre-registration step may be achieved if the
setup joint is equipped with joint sensors such as encoders. In
another exemplary configuration, the tip of a sensor equipped tool
may be used to touch or register the known feature on the table 12
to locate the table 12 with respect to other equipment within the
surgical suite.
[0062] The kinematics of the RCM 22 can also be calculated by
holding the tip of a fiber optic equipped tool in an arbitrary
fixed location and cycling through the various axes of the RCM 22.
By keeping the tip in a fixed location, the relative changes to the
fiber origin can be observed, and thus the kinematics of the system
can be determined and localized to the table 12. Once localized to
the table reference frame TRF, the tool may then be localized to
the fluoroscopy reference frame FF, as discussed above.
[0063] In addition to adding the sensors discussed in the above
techniques, additional modifications may be made to the location of
the fiber base to facilitate registering the fiber sensor to the
physical structure within the suite, such as, for example, the
table 12. For example, one modification is to extend the length of
a fiber in the catheter so that the origin/base can be extended out
of the splayer 24 and attached to a fixture having a known location
on the table 12. Once localized to the table reference frame TRF,
the tool may then be localized to the fluoroscopy reference frame
FF, as discussed above.
Registration to a 3D Model
[0064] Registration of the tool to a 3D Model is also contemplated
in this disclosure. Such registration may be performed directly
from the fiber sensor reference frame FRF to the 3D Model frame
AMF. In one exemplary technique, the operator is utilized. When the
tool (such as the catheter) is inserted into the patient,
tortuosity can be visualized from the fiber sensor data, as well as
on the pre-operative 3D Model. To register the tool in the 3D
Model, the operator may translate and rotate the 3D Model so that
distinct images and/or features in the tortuosity match or line up
with the shape of the fibers. However, in using this technique,
every time the patient moves, the tool should be re-registered.
[0065] In another exemplary arrangement, rather than having an
operator manually match features in the tortuosity, in one
technique, a computer algorithm such as automated geometric search
or mathematical optimization techniques that segments the model and
matches the model and tool shape dynamically may also be used to
match various shapes or features from the fiber sensor to the 3D
pre-operative Model. However, if the patient moves, even slightly,
the 3D Model could be mis-registered. Thus, the algorithms may be
used to re-register the tool automatically or the user could use an
input device, such as a track ball or mouse to move the 3D Model
manually.
[0066] Another proposed technique may be used to register the fiber
sensor to the 3D Model through the fluoroscopy image, as
illustrated in FIG. 3. In this technique, any of the above
described techniques for registering the surgical tool 12 to the
fluoroscopy reference frame FF may be utilized. To register the
fluoroscopy reference frame FF to the 3D Model reference frame AMF,
in one exemplary configuration, specific anatomical landmarks may
be used to provide recognizable reference points. The only
requirement for this approach is to have an anatomical landmark
that is recognizable in both the fluoroscopy reference frame FF, as
well as the pre-operative 3D Model reference frame AMF. Once the
recognizable point is identified in the fluoroscopy image, the 3D
Model may then be rotated by the operator to line up the recognized
points in the fluoroscopy images with the 3D Model images. This
action serves to register the fluoroscopy reference frame FF to the
frame of the 3D Model AMF. As the tool has previously been
localized to the fluoroscopy reference plane FF, so once the
fluoroscopy reference plane FF is registered, the tool's location
within the patient's anatomy may be determined with reference to
the 3D Model localizing the tool to the 3D Model. In one exemplary
configuration, a visual representation of the tool, based on the
transformation matrix, may be displayed on the 3D Model. In this
manner, the tool operator may then navigate the tool through the 3D
Model.
[0067] While certain of the above described techniques utilized
distinct marked points of a tool, such as a medical catheter, to
register the tool with the fluoroscopy image, it is also understood
that registration of the tool may occur based on the location of
the tool at the distinct anatomical landmarks. In other words, as
the tip of the tool can be driven to a known anatomical location in
the patient, the 3D Model may then be rotated by the user to
overlay the known anatomical location in the 3D Model with the
fluoroscopy image, in which the known anatomical location is
visible. Such action will also serve to register the tool with the
3D Model or localize the tool in the reference frame of the 3D
model reference frame AMF.
[0068] In another exemplary configuration, instead of, or in
addition to, having the user manually rotate the 3D Model to
correspond with the fluoroscopy image to line up distinct landmarks
visible in both the fluoroscopy image and the 3D Model, the
computer 36 may be programmed to employ a suitable algorithm such
as automated geometric search or mathematical optimization
techniques configured to match a distinct shape measured by the
fiber sensor with a corresponding shape in the 3D Model. In this
manner, the tool may also be registered with the 3D Model. The
accuracy of this method will depend on the size of vessel that the
tool is in, and the degree of curvature of the tool. Accuracy will
be improved if the tool is in a smaller vessel and will be worse if
the tool is in larger vessels. This automated technique can also be
used in conjunction with the manual techniques described above. For
example, the computer may be programmed to do automatic
registration and suggest a preferred registration but the user may
do final adjustments of the model. Once the tool is localized in
the 3D Model of the patient's anatomy, the user may then proceed to
maneuver the tool in the 3D Model.
[0069] Another technique that may be utilized to register the tool
to the 3D Model through fluoroscopy system 30 involves the use of
radiopaque markers. More specifically, radiopaque markers can be
fixed to the anatomy. However, these markers would need to be
present during pre-operative imaging when the 3D Model is created,
and remain in the same location during intraoperative fluoroscopy
imaging. With this technique, the position of these markers in the
fluoroscopy reference frame FF can be used to correlate to the same
markers in the 3D Model reference frame AMF, thereby registering
the fiber sensor to the 3D Model reference frame AMF.
Three-dimensional angiography may also make it easier to register
the tool to the 3D Model by facilitating acquisition of the model
in realtime, i.e. while the patient is on bed. While the model
might drift away from the real anatomy when the operation is
carried out, it may be advantageous to obtain the model and perform
operations in the same spot.
[0070] Another technique that may be utilized to register the
surgical tool to a 3D Model utilizes intravascular imaging. This
technique allows for 3D visualization of a surgical tool, such as,
a catheter, in the anatomy, but without the use of fluoroscopic
imaging. Such a technique can benefit both physicians and patients
by improving the ease of tool navigation, as well as and reducing
radiation exposure of personnel inside the operating room.
[0071] Turning now to FIG. 8, the registration technique 600 may
begin by utilizing a sensor 602 operatively coupled to the tool to
sense a shape of the tool 604 while in the patient. This sensed
shape is then mathematically correlated against features of the
vascular model such as centerlines or walls in which a larger
correlation value corresponds to a better match. The correlation
can be performed in real-time on each shape or by batch processing
a sequence of shapes. This proposed technique assumes that the tool
will always follow a unique configuration through the vasculature,
and thus, a global maximum for the correlation exists. However, the
correlation may return many local maxima since the tool
configuration may follow many different paths between fixed distal
and proximal ends. Choosing an incorrect maximum introduces
registration error. Furthermore, in some cases, the pre-operative
3D model may differ from the actual vasculature for a number of
reasons, including, for example, patient motion or inaccuracies in
pre-operative sensing. Such situations also may lead to
registration error.
[0072] Recent advances in intravascular imaging technology have
brought about sensors 604 that can provide information about the
local structure of vessel walls 606. Such information may be used
for shape registration and environmental mapping. Two examples of
these sensors are intravascular ultrasound (IVUS) probes, and
optical coherence tomography (OCT). Intravascular ultrasound
periodically produces a 2-D cross-sectional view of the blood
vessel either to the sides of the catheter in standard IVUS or to
the front of a catheter in Forward-Facing IVUS. Optical Coherence
Tomography periodically produces a local 3D view of the vessel into
which the tool is inserted. The images produced by these
technologies may be processed to provide an estimate of a curve or
surface representing the vessel wall 606. The sensors 604 may also
determine the location of the catheter's endpoint within the
vascular cross-section. Use of the sensors coupled with the tool
602 to provide shape information coupled with information
obtainable from sensors 604 configured to provide information about
the vessel walls 606 can assist in defining the 3D shape of the
blood vessel 608.
[0073] Once the shape of the vessel is defined or otherwise
reconstructed using the combined sensor data, the shape can be
mathematically correlated to the 3D model 610, thereby registering
the tool to the 3D Model 612. In implementation, the 3D
reconstruction and correlation steps may be combined into a single
recursive filtering algorithm. A Bayesian filter (e.g. Extended
Kalman Filter (EKF), Unscented Kalman Filter (UKF), or Particle
Filter) may be used to develop an estimate of the tool's position
relative to the pre-op 3D model given both imaging and sensor 602
information. The filter's state is a set of points or a parametric
curve representing the position and shape of the tool 602 with
respect to the pre-op 3D model, as well as the rate of change of
this shape. For accurate registration, patient motion may also be
taken into account. Thus, the filter's state may also contain
warping parameters for the pre-op 3D model. These warping
parameters may be evenly distributed, or may be selected based on
the structure of the anatomy around the vasculature. The motion of
the structure of the anatomy around the vasculature may be measured
using visible light tracking technologies such as stereoscopic
cameras, structured light tracking technologies, and/or other
localization technologies attached to the patient skin.
[0074] The recursive filtering algorithm operates by predicting the
motion of the tool in the 3D model, then performing an update of
the filter hypothesis given new sensor measurements. At each
time-step, a kinematic model of the catheter and control inputs
such as current pull-wire tension and displacement may be used to
perform the filter's motion update. The filter's measurement update
may apply a correction to the tool registration and model warping
parameters by comparing a predicted vessel wall with the sensed
position and orientation of the vessel from the imaging and sensor
measurements. The update effectively executes the correlation
between 3-D sensor information and the 3D model. Performing these
correlations repeatedly in a recursive filtering framework may
provide a real-time catheter position estimate. Furthermore, the
filter's parameters may be tuned such that differences between the
measurements and the model over a small time constant (ms) will
lead to changes in the catheter position estimate in order to
filter out high-frequency sensor noise. Differences over a large
time constant (seconds) may lead to changes in the model's warping
parameters.
[0075] Thus, once the tool has been registered to the 3D model, the
location of the tool within the 3D model may be determined,
allowing an operator to drive the tool within the vasculature using
the 3D model without requiring intra-operative fluoroscopy.
[0076] Sensors 604 may also be utilized to sense the environment
around the tool. Thus, once the tool is registered to the 3D model,
this environmental information, such as, for example, vascular
occlusions may be displayed at the correct position in the 3D
Model.
[0077] More specifically, after tool registration, the
intravascular imaging sensor 604 provides a mechanism to sense and
display features of the environment surrounding the tool without
the use of fluoroscopy. There are many ways to display this
information. One non-limiting option is to simply provide a display
of a real-time view of the imaging sensor's output alongside a view
of the catheter's location in the 3D model or superimposed on top
of the 3D model. Another option may be to analyze the intravascular
image to detect environmental changes. For example, IVUS image
processing techniques can be used to detect areas of plaque in the
image. This information can be used to annotate the IVUS image in
order to alert the physician to environmental conditions. Since a
combination of IVUS and sensor data 602 may provide 3D information
on the structure of these plaque formations, the 3D pre-op model
can also be annotated. In this way, the existing work that has used
IVUS to perform vascular sensing may be leveraged by the combined
IVUS and sensor system to provide a 3D view of the environment to
the physician.
Exemplary Error Reduction Methods
[0078] Turning now to FIGS. 9-11, exemplary systems and methods for
reducing measurement error, e.g., of an incremental measurement
sensor (IMS), is described in further detail. Registration, as
described above, is generally the mathematical process of computing
the orientation and position of all or portions of a shape
(sequence of points) in some coordinate frame. Registration can fix
the orientation and position of the fiber at one or more points
(for six degrees of freedom of constraints at each point) or
registration can be carried out with fewer constraints at a single
point or multiple points. Inherently, though, registration is the
method of positioning and orienting a sensor frame in a reference
frame.
[0079] For any IMS, one type of registration is to determine the
position and orientation of the proximal end ("origin" of the IMS)
in the fluoroscopy, or world, coordinate frame and display that to
the user in the virtual environment. By determining the
transformation of the "origin" of the IMS to the world coordinate
system, the points in the IMS can be transformed and displayed in
the world coordinate system throughout the procedure. However, the
accuracy of this registration procedure can be difficult and
inadequate for applications that require sub-mm accuracy at the
distal end of the sensor. Inaccuracy will result if the origin of
the IMS moves; if the origin does move, it needs to be tracked in
the world coordinate frame, which may lead to error. In addition,
the error inherent in the sensor could cause errors in the sensor
measurement. For instance, in a fiber optic shape sensor, the
tighter the bends, the less accurate the sensor is. Tight bends
that may occur early in the path of the fiber either due to
mechanical structures in the device or pathways in the anatomy,
cause additional error in the orientation measurement. This
orientation error may be magnified by the lever arm of the fiber
causing the tip to be inaccurate by centimeters. This is unsuitable
for many applications since sub-millimeter accuracy is desired.
Additional Registration Points
[0080] In one exemplary approach illustrated in FIG. 9, multiple
positions of registration along the length of an elongate member
900 may be provided to reduce the error in a sensor 902. In one
example, the sensor 902 is an IMS fiber sensor configured to output
a position of the elongate member at the sensor 902. The elongate
member, e.g., a catheter, may be generally fixed at a proximal end
904 such that the position of the proximal end 904 may generally be
known. A position of the elongate member 900 is known at one
additional point 906, which is proximal to the distal portion of
the sensor 902. While only one additional point 906 is illustrated,
any number of additional points proximal to the sensor 902 may be
used, as described further below. Using the additional piece(s) of
information at the point(s) proximal to the sensor 902 may be used
to register the IMS at a position distal of the "origin," i.e., at
the sensor 902, helping to reduce orientation error propagated from
the proximal end to a position error in the distal end. The nature
of the fiber and catheter is that the proximal end is the only
place that the fiber is attached to the system, but there are a
number of ways of acquiring registration along the length of the
fiber.
[0081] Registration of a shape or series of points of the IMS
sensor 902 can be used to improve registration. This can be
performed by acquiring 3D shapes or obtaining spatial information
from 2D images, i.e., of the elongate member 900. For example, a
plurality of points 906, 908, 910, and 912 may be employed to
determine a shape of the elongate member 900 at each of the points
906, 908, 910, and 912, such that a shape or position of the
elongate member 900 is known. There are many sources of a
three-dimensional anatomic model in robotic catheter procedures
that may be used in this manner. For instance, models could be
generated from a rotational angiography, computed tomography scan,
or other standard imaging technique. Also, any known shape that the
catheter passes through could also contribute to registration just
as in a 3-D model, such as an engineered semitortuous introducer, a
curve in the take-off of the sheath splayer, an anti-buckling
device feature, etc. In all these cases, the catheter and IMS will
be passing through a known shape that can be used to register the
position and orientation of a distal portion of the catheter.
[0082] Once the catheter is known to pass through a given
three-dimensional shape defined by the points 906, 908, 910, and
912, providing the registration is an optimization problem to find
the most likely position and orientation of the sensed shape in
relation to the model of the elongate member 900. This solution
could be completed using many standard computer graphics matching
techniques or posed as a numeric optimization problem.
[0083] Another example of other 3D information may be to add other
localization sensors such as an electromagnetic sensor to various
critical points along the catheter. More specifically, at least one
of the proximal points 906, 908, 910, and/or 912 may have an
electromagnetic sensor. The generally absolute measurements of
position at these locations can reduce any error propagated from
the fiber at a portion distal of the position of any
electromagnetic sensors at points 906, 908, 910, and/or 912.
[0084] Another exemplary method of obtaining additional shape
information is to use computer vision techniques on the fluoroscopy
images to track the catheter and then feed that information into
the system to improve registration. Because two-dimensional imaging
will provide less information than a 3D model it may not provide as
much information to reduce error, but it would likely remove error
at least in the plane of the image. This registration might also
not be needed constantly during a procedure, but may be used when
imaging, e.g., fluoroscopy, is active and the operator wants more
accuracy of the tracked catheter. Accordingly, a position of a
proximal portion of the elongate member 900, e.g., along points
906, 908, 910, and/or 912 may be visualized in 2D and used to
increase accuracy of measurements at the sensor 902.
[0085] In a way, the above exemplary approaches to optimizing
position measurements of the sensor 902 provide an alternate
registration location to a proximal portion of the elongate member,
e.g., at the base of the fiber at the splayer attached to the RCM.
However, in these examples registering a shape to a 3D model does
not necessarily completely replace the registration at the splayer
(not shown in FIG. 9). Using an optimization technique that finds
the most likely position and orientation of the sensor given both
the anatomical registration and the distal splayer registration is
probably the best way to reduce overall error and achieve a strong
overall pose of the catheter in relation to the anatomy. This
algorithm can also take advantage of any proximal motion of the
catheter, such as shaft insertion, etc. However, this algorithm can
be time consuming and computationally intensive, especially when a
good starting point is not given. The registration of the origin of
the IMS could be used as such a starting point.
[0086] Another exemplary approach in the case where global
registration of position is not needed would be to allow the user
to designate a specific position on the catheter that is
constrained laterally, e.g., by the anatomy of a patient. For
example, as shown in FIG. 10 an elongate member 1000 is partially
inserted into a patient 1100, who remains generally fixed in
position on a table 1102. In this manner, a distal portion 1004 of
the elongate member 1000, which includes an IMS 1006 for measuring
position of the elongate member 1000, is received within the
patient 1100, while a proximal portion 1002 remains outside the
body of the patient 1100. The insertion site 1008 of the elongate
member 1000 into the patient 1100 is known and is substantially
fixed, such that a position of the elongate member 1000 is known at
the insertion site 1008 in at least two dimensions. The insertion
site 1008 position, at least in two degrees of freedom, could be
registered in addition to the splayer 1010, while the insertion of
the elongate member 1000 is updated accordingly. With this
technique, the orientation of the elongate member 1000 and/or a
measurement fiber at this designated `base` (i.e., the insertion
site 1008) of the shorter effective fiber would depend on the
actual orientation of the fiber (including twist) and the insertion
would depend on the commanded fiber insertion. However, the shape
distal to the insertion site 1008 could be updated rapidly and
without any error from the shape of the proximal portion 1002,
which is proximal to the designated base position, i.e., the
insertion site 1008. Accordingly, error in position measurements of
the elongate member 1000 anywhere along the distal portion 1004 is
greatly reduced by effectively reducing the measurement length of
the IMS 1006.
[0087] The above exemplary approach may be particularly
advantageous for relative position display (essentially a
coordinate frame not correctly registered to the
fluoroscopy/anatomical coordinate frames) but may ultimately make
it difficult to determine a global position of the tip of the
catheter. However, in many applications that do not superimpose the
catheter directly over a model of anatomy (or imply perfect
registration) it would be perfectly appropriate.
Selective Filtering
[0088] Other exemplary approaches are generally based around the
idea that filtering is a method of sacrificing responsiveness in
the time domain to reduce the shape and position error. For
example, an operator of an elongate member may generally be
focusing on the distal end of the catheter during a procedure, and
the proximal end will generally be stable and not moving quickly.
As a result, selective filtering on proximal shape can be applied
to reduce orientation error in the proximal region while the distal
region does not filter the signal. More specifically, in the
example shown in FIG. 9, position data associated with the distal
region of the elongate member 900 between the points 912 and 902 is
not filtered, while the position data of a proximal region between
points 906 and 912 is filtered. The selective filtering of the
proximal region reduces the influence of proximal error by
filtering it out (for instance, averaging over multiple time steps)
while, the distal portion will be updated as fast as possible using
the orientation and position of the proximal section as a stable
base.
[0089] In one exemplary illustration, one could average the
positions and orientation over time in the global frame, or you
could average incremental changes in orientation over time and then
integrate the averaged orientation changes to yield the final
shape. Integrating the incremental changes could be a more accurate
average particularly in the case of a fiber-based measurement
system, e.g., FOSSL, because the actual measurement is being
determined from an indication of strain, which is proportional to
bend angle and orientation. Thus, white noise may be present on the
level of the strain/incremental orientation as opposed to the final
incremented position and orientation.
[0090] Because the attention of the operator will likely be focused
on the distal tip or portion, the result of proximal filtering will
be a responsive system that exhibits less error. The exact location
where filtering starts or stops may vary based on the application
and it is even possible to apply a variable level of filtering
along the fiber with the maximum filtering at the proximal end,
e.g., between points 906 and 912, and the minimum at the distal
end, e.g., distal of the point 902. In one case when absolute
registration is not needed, it is simple to ignore the proximal
portion(s) of the fiber and treat the distal portion of the fiber
as a non-grounded shape sensor that can be oriented
arbitrarily.
[0091] In examples where the proximal end is filtered heavily in
relation to the distal end, it may be useful to detect when the
proximal shape changes significantly in a short period of time,
such as a prolapsed catheter at the iliac bifurcation (a rapid
motion of the shaft of the catheter bulging up into the aorta).
This can be accomplished by noting when the new catheter shape is
significantly different than the filtered shape. In this case, the
filtering on the proximal end can be reduced so that the operator
sees the most up to date information and can react accordingly.
After the proximal shape remains more constant for a period of
time, the filtering can increase, again reducing error at the tip.
To prevent sudden jumps in the rendered data, temporarily-variable
filtering algorithms can be implemented in such a way as to provide
continuity of the filter output, which may increase the appearance
of smoothness and reduce noise of the measurement.
[0092] An anatomical model need not be used to separate distal and
proximal updates of an IMS localization technology. One exemplary
approach would be to apply relatively heavy filtering on the
proximal end of the fiber, e.g., between points 906 and 912, and
light filtering on the distal end, e.g., distal of point 902, as
described above. Specific aspects of this sensor when used in a
robotic system could modified, such as updating insertion
immediately without filtering since axially the catheter is
relatively stiff and low in error (assuming the catheter does not
buckle). This is potentially problematic because by filtering
different dimensions at different rates, trajectories can become
skewed, producing measurement points that do not lie along the true
trajectory of the device. Additionally, filtering could be
accelerated when the proximal measurement changes significantly in
relation to the filtered version to give more responsive feedback
during a prolapsed situation.
[0093] It should be noted that instinctiveness computations are
often computed from orientations propagated from the registered
base of the fiber. Instinctiveness refers generally to matching and
orientation or a location of a device such as a catheter with a
visualization device that is used to control the catheter, such as
an image of the catheter. Since instinctiveness includes absolute
orientation measurements, they may be computed separate from any
intermediate registration or clipping techniques. On the flip side,
because distance does not magnify orientation errors, there is
little or no extra error from the distance from the base to the tip
of the fiber in instinctiveness measurements. Furthermore,
instinctiveness measurements generally do not require a fast update
rate so it is possible to filter heavily to reduce orientation
error and noise.
[0094] Turning now to FIG. 11, an exemplary process 999 for
determining a position of a distal portion of an elongate member is
illustrated. Process 999 may begin at block 1110, where an
incremental measurement sensor may be provided. For example, as
described above an elongated member 900 may have an IMS 902
positioned along a distal portion of the elongate member 900. In
some exemplary approaches, the elongate member 900 includes a
fiber, which may be employed to determine position and/or
orientation data of the elongate member 900. Process 999 may then
proceed to block 1112.
[0095] At block 1112, registration data may be applied at one or
more proximal locations along the elongated member. For example, as
described above a position of the elongate member 900 at any one or
more of points 906, 908, 910, and/or 912 may be registered. For
example, a proximal position of one or more proximal locations may
be used to increase accuracy of measured data from the IMS 902.
[0096] In some examples, one of the proximal locations used to
apply registration data along the elongated member 900 includes a
proximal attachment of the elongated member 900, e.g., at a
proximal end 904. The proximal attachment at the proximal end 904
may general fix a portion of the elongated member at the first one
of the locations. In other examples, a position of the proximal
location(s) may be determined using an electromagnetic sensor, or
by using a two-dimensional image of the one or more proximal
locations, e.g., as obtained by fluoroscopy during a procedure. In
still another example, applying the registration data at the one or
more proximal locations may include constraining a lateral position
of the one or more proximal locations. Merely as one example, as
described above an insertion site 1008 associated with a patient
1100 may generally constrain an elongate member 1000 laterally at
the insertion site 1008. Accordingly, the generally fixed insertion
site 1008 indicates a position of the elongate member 1000 in at
least two dimensions at the insertion site 1008.
[0097] Proceeding to block 1114, a proximal signal of the elongated
member may be selectively filtered, e.g., in relation to a distal
signal of the elongated member. Filtering of the proximal signal
may occur at a different rate than the filtering of the distal
signal. In one example, heavier or more intrusive filtering of the
proximal signal may be employed, especially during a procedure
where the proximal portion(s) of the elongate member do not change
rapidly. In some cases, filtering may include averaging an
incremental orientation position change in the proximal signal.
Moreover, variable filtering methodologies may be used. For
example, as described above a heavier filtering methodology may be
employed only during such times that a position of the proximal
portion of the elongate member is not rapidly changing position.
Upon detection of a rapid or unexpected change, the filtering of
the proximal portion may cease or be reduced. Filtering at the
initially heavier setting may resume after a predetermined period
of time expires, in which the proximal portion of the elongate
member maintains a same position or does not rapidly change
position during such time. Process 999 may then proceed to block
1116.
[0098] At block 1116, a position of the incremental measurement
sensor may be determined. For example, as described above a
position of an IMS 902 may be determined based at least upon the
registration data from the one or more proximal locations. In this
manner, measurement error may be reduced since less incremental
error occurs over the length of the elongate member. Alternatively,
or in addition, a position of the incremental measurement sensor
may be determined based at least upon a filtered proximal signal.
In such cases, selectively filtering may reduce a fluctuation of
the determined position of incremental measurement sensor by
generally smoothing out position signals relating to the proximal
portion(s) of the elongate member.
CONCLUSION
[0099] The methods described above are the examples of registration
using known data about the location of the catheter in relation to
the anatomy. The first example does not include extra localization
information while the second example assumes some knowledge of the
shape of the anatomy or other features along the path of the
catheter. Between these examples, there are a number of other ways
to get partial registration at one or more points along the fiber
to reduce orientation error propagated down the fiber.
[0100] The exemplary systems and components described herein,
including the various exemplary user interface devices, may include
a computer or a computer readable storage medium implementing the
operation of drive and implementing the various methods and
processes described herein. In general, computing systems and/or
devices, such as user input devices included in the workstation 31
or any components thereof, merely as examples, may employ any of a
number of computer operating systems, including, but by no means
limited to, versions and/or varieties of the Microsoft Windows.RTM.
operating system, the Unix operating system (e.g., the Solaris.RTM.
operating system distributed by Oracle Corporation of Redwood
Shores, Calif.), the AIX UNIX operating system distributed by
International Business Machines of Armonk, N.Y., the Linux
operating system, the Mac OS X and iOS operating systems
distributed by Apple Inc. of Cupertino, Calif., and the Android
operating system developed by the Open Handset Alliance.
[0101] Computing devices generally include computer-executable
instructions, where the instructions may be executable by one or
more computing devices such as those listed above.
Computer-executable instructions may be compiled or interpreted
from computer programs created using a variety of programming
languages and/or technologies, including, without limitation, and
either alone or in combination, Java.TM., C, C++, Visual Basic,
Java Script, Perl, etc. In general, a processor (e.g., a
microprocessor) receives instructions, e.g., from a memory, a
computer-readable medium, etc., and executes these instructions,
thereby performing one or more processes, including one or more of
the processes described herein. Such instructions and other data
may be stored and transmitted using a variety of computer-readable
media.
[0102] A computer-readable medium (also referred to as a
processor-readable medium) includes any non-transitory (e.g.,
tangible) medium that participates in providing data (e.g.,
instructions) that may be read by a computer (e.g., by a processor
of a computer). Such a medium may take many forms, including, but
not limited to, non-volatile media and volatile media. Non-volatile
media may include, for example, optical or magnetic disks and other
persistent memory. Volatile media may include, for example, dynamic
random access memory (DRAM), which typically constitutes a main
memory. Such instructions may be transmitted by one or more
transmission media, including coaxial cables, copper wire and fiber
optics, including the wires that comprise a system bus coupled to a
processor of a computer. Common forms of computer-readable media
include, for example, a floppy disk, a flexible disk, hard disk,
magnetic tape, any other magnetic medium, a CD-ROM, DVD, any other
optical medium, punch cards, paper tape, any other physical medium
with patterns of holes, a RAM, a PROM, an EPROM, a FLASH-EEPROM,
any other memory chip or cartridge, or any other medium from which
a computer can read.
[0103] Databases, data repositories or other data stores described
herein may include various kinds of mechanisms for storing,
accessing, and retrieving various kinds of data, including a
hierarchical database, a set of files in a file system, an
application database in a proprietary format, a relational database
management system (RDBMS), etc. Each such data store is generally
included within a computing device employing a computer operating
system such as one of those mentioned above, and are accessed via a
network in any one or more of a variety of manners. A file system
may be accessible from a computer operating system, and may include
files stored in various formats. An RDBMS generally employs the
Structured Query Language (SQL) in addition to a language for
creating, storing, editing, and executing stored procedures, such
as the PL/SQL language mentioned above.
[0104] In some examples, system elements may be implemented as
computer-readable instructions (e.g., software) on one or more
computing devices (e.g., servers, personal computers, etc.), stored
on computer readable media associated therewith (e.g., disks,
memories, etc.). A computer program product may comprise such
instructions stored on computer readable media for carrying out the
functions described herein.
[0105] With regard to the processes, systems, methods, etc.
described herein, it should be understood that, although the steps
of such processes, etc. have been described as occurring according
to a certain ordered sequence, such processes could be practiced
with the described steps performed in an order other than the order
described herein. It further should be understood that certain
steps could be performed simultaneously, that other steps could be
added, or that certain steps described herein could be omitted. In
other words, the descriptions of processes herein are provided for
the purpose of illustrating certain examples, and should in no way
be construed so as to limit the claims.
[0106] Accordingly, it is to be understood that the above
description is intended to be illustrative and not restrictive.
Many examples and applications other than the examples provided
would be apparent upon reading the above description. The scope
should be determined, not with reference to the above description,
but should instead be determined with reference to the appended
claims, along with the full scope of equivalents to which such
claims are entitled. It is anticipated and intended that future
developments will occur in the technologies discussed herein, and
that the disclosed systems and methods will be incorporated into
such future examples. In sum, it should be understood that the
application is capable of modification and variation.
[0107] All terms used in the claims are intended to be given their
broadest reasonable constructions and their ordinary meanings as
understood by those knowledgeable in the technologies described
herein unless an explicit indication to the contrary in made
herein. In particular, use of the singular articles such as "a,"
"the," "said," etc. should be read to recite one or more of the
indicated elements unless a claim recites an explicit limitation to
the contrary.
[0108] The Abstract of the Disclosure is provided to allow the
reader to quickly ascertain the nature of the technical disclosure.
It is submitted with the understanding that it will not be used to
interpret or limit the scope or meaning of the claims. In addition,
in the foregoing Detailed Description, it can be seen that various
features are grouped together in various examples for the purpose
of streamlining the disclosure. This method of disclosure is not to
be interpreted as reflecting an intention that the claimed
embodiments require more features than are expressly recited in
each claim. Rather, as the following claims reflect, inventive
subject matter lies in less than all features of a single disclosed
embodiment. Thus the following claims are hereby incorporated into
the Detailed Description, with each claim standing on its own as a
separately claimed subject matter.
* * * * *