JP2010510815A - Adaptive navigation technology for navigating a catheter through a body passage or cavity - Google Patents

Abstract

A method of using an assembled 3D image to build a 3D model for determining a route through a luminal network to a target. The 3D model is automatically registered to the actual position of the probe by tracking and recording the position of the probe and by continuously adjusting the registration between the model and the display of the position of the probe. The The registration algorithm becomes dynamic (elastic) as the probe approaches a smaller lumen at the periphery of the network where movement has a greater effect on registration between the model and the representation of the probe.

Description

(Related application)
This application is U.S. Provisional Patent Application No. 60 / 865,379, filed on November 10, 2006, under the name Adaptive Navigation Method, and is named Adaptive Navigating Technology for Cathre Thorough A6. US Provisional Patent Application No. 60 / 867,428, filed on November 28, and Adaptive Navigation Technique for Navigating A Catheter Through A Body Channel Or Cavity, filed February 1, 2007 Claiming priority to Application No. 60 / 887,663, the above application All are incorporated by reference herein.

(Background of the present invention)
Breakthrough techniques have emerged that allow navigation of the catheter tip to a predetermined target through tortuous passages such as those found in the pulmonary artery system. This technique compares the real-time movement of a placeable guide (LG) against a three-dimensional digital map of a targeted area of the body (the person skilled in the art will give some examples It will be appreciated that, for example, the pulmonary pulmonary airway will be used below for purposes of illustration, although it may be used in any body cavity or organ such as the circulatory system, digestive system, pulmonary artery system, etc.

  Such technologies are all described in Patent Document 1, Patent Document 2, Patent Document 3, Patent Document 4, Patent Document 5, Patent Document 6, Patent Document 7, Patent Document 8, Patent Document 9, Patent Document 9, which belong to Gilboa or Gilboa et al. Patent Document 10, Patent Document 11, and Patent Document 12, Patent Document 13, Patent Document 14, and Patent Document 15 belonging to either Gilboa or Gilboa et al. All of these references are incorporated herein in their entirety.

US Pat. No. 6,188,355 US Pat. No. 6,226,543 US Pat. No. 6,558,333 US Pat. No. 6,574,498 US Pat. No. 6,593,884 US Pat. No. 6,615,155 US Pat. No. 6,702,780 US Pat. No. 6,711,429 US Pat. No. 6,833,814 US Pat. No. 6,974,788 US Pat. No. 6,996,430 US Patent Application Publication No. 2002/0193686 US Patent Application Publication No. 2003/0074011 US Patent Application Publication No. 2003/0216639 US Patent Application Publication No. 2004/0249267

  One aspect of this background art collectively relates to the registration of CT images used as a three-dimensional digital map for the actual movement of LG through the pulmonary artery system. The user interface shows three separate CT-based images that are reconstructed by software from the x, y, and z directions, as well as the position of the LG superimposed at the intersection of the reconstructed images. If the CT image does not accurately reflect the actual location of the airway, as the LG is advanced, the LG will appear to divert quickly out of the airway, thereby reducing the usefulness of the navigation system.

  Currently, registration points at selected known landmarks in the central area of the lung are used to register or align a CT-based digital map with the patient's chest cavity. These registration points are initially chosen during the planning phase and marked on the inner lung surface. At the beginning of the procedure, the corresponding points are touched and recorded using LG assisted by a bronchoscope in the patient's airway. Doing so allows the computer to align the digital map with the data received from the LG so that an accurate representation of the location of the LG is displayed on the monitor.

  However, due to various factors, registration accuracy decreases as the distance between the LG and the registration point increases. In other words, the navigation system becomes less accurate at the periphery of the lung where accuracy is most needed. This is due to a variety of factors, two of which are the focus of the present invention. The first factor relates to the rigidity of CT digital images used as digital maps by current systems, while the lung structure is flexible. Second, errors are synthesized as the distance from the last registration point increases. The synthesized error combines with the flexible airway to become LG that appears to be outside the airway on the CT image.

  As a result of the accumulated inaccuracies, the performance of existing systems is limited. For example, once the bronchoscope is too large to proceed, the existing system ignores the flexibility and inaccuracies created by the movements inside the living airway and whether the LG is advanced in the direction of the target Provide guidance to users regarding In addition, guidance instructions to the target are provided without considering the airway geometry leading to the target. As a result, the user gradually advances LG and watches whether LG is moving in the direction of the target. Otherwise, the LG is retracted and the user “feels” another airway that can lead to the target rather than looking directly at the target on the CT cross section. Therefore, two problems arise. First, the LG no longer appears to be located in the airway. Second, the provided guidance does not guide the user along a logical path, it merely provides an approximate direction to the lesion.

  The present invention solves these two problems by using a unique algorithm to create a BT skeleton, a three-dimensional virtual map of the bronchial airway, and by continuously and adaptively matching the LG pathway to the BT skeleton. Work on. Due to the increased accuracy of the BT skeleton and registration, three-dimensional guidance is extended beyond the limits of bronchoscopes.

FIG. 1 is a diagrammatic representation of the algorithm of the present invention. FIG. 2 is a table showing the relative importance of the sources of error uncertainty encountered in the method of the present invention. FIG. 3 is a chart showing the steps of the route generation process of the present invention. 4 to 13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. 4-13 are screen captures of the user interface during the path generation process of the present invention. FIG. 14 is a screen capture of the user interface of the present invention while the procedure is being performed on a patient.

(Detailed Description of the Invention)
(How to generate a route to a target in the lung)
The present invention includes a unique method of generating a BT skeleton so that an accurate and logical path to the target can be formed. In general, the method begins with an algorithm that automatically detects the trachea inside the CT volume, which is a three-dimensional image made from multiple CT scans, and uses this as a starting point for BT generation. Different segmentation steps are then applied to mark those voxels of the CT scan representing air in the bronchi. The segmented and filtered data is then skeletonized, ie, the sensed airway centerline is defined and used to build an anatomically valid virtual model of the airway.

  More specifically, a method for generating a route to a target in the lung is as follows.

1. Bronchial tree generation Bronchial tree generation is a fully automatic process that runs in the background and is therefore transparent to the user while working with application software.

2. Automatic Seed Point Detection Automatic seed point detection is an algorithm that detects the airway by looking for a tubular object having a density of air in the upper region of the CT volume. The center of gravity of the found tubular object is defined as a seed point for further segmentation.

3. Segmentation: Lung Air Identification Segmentation is described in the Region Growth Algorithm (Handbook of Medical Imaging, Processing and Analysis, Isaac N. Bankman, Academic Press, 2000, page 73, incorporated herein by reference in its entirety. The algorithm defines and displays bronchial airways from CT volume images of the human chest. The purpose of the region growing algorithm is connected to the seed point that is the starting point, and satisfies the following condition: the Hounsfield value (HU) at all of these points is below a predefined maximum threshold It is to create a uniform region of points.

  The implemented process is fully automated, iterative and consists of several steps:

3.1 Segmentation of anatomical features:
Its purpose is to mark (or segment) the portion of the entity's voxel (voxel = Volume Pixel) within the recognizable features of the luminal network. This recognizable feature is used as a starting point. For example, in the case of a pulmonary airway that is a luminal network, preferably the trachea is selected as an anatomical feature. As a result, portions of voxels representing air in the trachea are marked while avoiding noise and artifact-induced bubbles in the CT image. For this purpose, region growth with a high threshold is then applied in the volume. For example, when blood vessels constitute a lumen network, the aorta can be used as an anatomical feature.

3.2 Adaptive threshold detection for region growing algorithms:
3.2.1 From step 3.1, starting at the boundary of the previously segmented area, a number of iterations of the region growing algorithm are performed. For each iteration, the following steps occur:

  3.2.1.1 A threshold is defined and all voxels below the threshold are considered to contain only air and are therefore segmented. This process is iterative, but the growth rate and geometry are not considered.

  3.2.1.2 After the segmentation process is completed for that iteration, the total number of segmented voxels inside the lungs that originate from the seed point is recorded.

  3.2.2 Next, the threshold is increased and the next iteration is performed. After this iteration is completed, the number of segmented voxels between the current iteration and the previous iteration is compared.

  3.2.3 Since the threshold increases with each iteration, each iteration can result in a larger number of combined voxels. Increasing the threshold means that more voxels are considered air.

  3.2.4 This event is considered a leak if the difference in the number of segmented voxels between two consecutive iterations has increased substantially. In fact, it means that where the bronchial wall is “broken” by segmentation and in addition to air in the lungs, the outer lung air is now connected to the segmented volume. means. Therefore, a conclusion can be drawn that the current threshold is too high and the threshold from the previous iteration is used.

  3.2.5 Finally, segmentation is performed with the selected threshold. This time, the segmentation result is added to step 1 and stored. This is used as a starting point for the next step.

3.3 Leakage control:
This is required to improve the results of the region growing algorithm with adaptive threshold local values by segmenting additional areas. The technique of section 3.2 is applied to all boundary points (points on the tissue) of the previously segmented area.

3.4 Propagation of geometric control waves:
This means that the paper "Hybrid Segmentation and Exploration of the Human Lungs," IEEE Visualization 2003, Dirk Bartz, Dirk Mayer, Jan Fischer, Sebastian Ley, Anxo del Ro, Stef Thust, Claus Peter Heussel, Hans-Ulrich Kauczor, and Wolfgang Described in Strasser, the entirety of which is incorporated herein by reference. This paper allows additional improvements over previous steps, using higher threshold levels due to the mechanism of geometric parameter control of growing branches.

3.5 “Template matching”:
This approach is based on the above paper by Bartz et al. And evaluates candidate areas below the template with indefinite density values (between -950 HU and -775 HU). This is organized in two stages. The first stage builds templates, which are used in the second stage to evaluate the local voxel neighborhood. First, 2D template matching applies 2D region growth starting from previously segmented boundary voxels. The threshold varies from a threshold above an indefinite density value interval (−775 HU) until the number of selected voxels is less than critical. Because it can be assumed that they did not leak. Based on the selected voxel area, a cyclic template of variable size is generated. In the second stage we apply 2D region growth. The shape of each combined segmented area is compared to the cyclic template set from the first stage. Positive comparison results are then selected and added to the segmentation.

3.6 Bubble filter:
Finally, a bubble filter is applied. Bubble filters are a combination of morphological dilation and erosion operations. It is used to remove small non-segmented areas (bubbles) from the last segmented area. These bubbles appear due to the noisy nature and artifacts of CT images.

4). Skeletonization and feature calculation Skeletonization and feature calculation involves extracting previously segmented bronchial centerlines, building an effective anatomical hierarchy of bronchial airways, calculating bronchial diameter and geometric features , Refers to the surface generation of each segmented bronchi. The following steps are involved:

4.1 Dilution algorithm:
The iterative object reduction technique is described in the paper A Sequential 3D Thinning Algorithm and Its Medical Applications, K'alm'an Pal'Agyi, Eric Sortin, Emesa Blog, Atmel. 17th Int. Conf. IPMI (2001) 409-415, the entire article of which is hereby incorporated by reference and converts previously segmented airways into a geometric skeleton representation. Used for.

4.2 Branch and node point detection:
A map of all skeleton voxels is generated, resulting in a list of neighboring voxels for each voxel. A voxel with three or more neighboring voxels is considered a “node point”. A voxel with two neighboring voxels is considered a point on the branch. The entire voxel map is rearranged as a graph with nodes and branches.

4.3 Fake branch filtering:
This is related to the following steps:
4.3.1 Identify and remove disconnected branches.

  4.3.2 Resolve the loop of the graph by removing the longest of the two branches connected to the common node.

  4.3.3 Remove relatively short leaves in the graph as a result of leakage.

  4.3.4 Remove leaves that are relatively close to each other.

4.4 Convert a graph to a tree:
The root point is found on the graph as the closest point to the seed point found in 1. The graph is converted into a binary tree. The branches are approximated by polynomials.

4.5 Labeling branches: logical and hierarchical:
This means that the paper Automated Nomenclature Labeling of the Bronchial Tree in 3D-CT Lung Images, Hiroko Kitaoka from Osaka University, Yongsup Park, Juerg Tschirren, Joseph Reinhardt, Milan Sonka, Goeffrey McLennan, and Eric A. Hoffman from University of Iowa, Lecture Notes in Computer Science, T.W. Dohi and R.D. Kikinis, Eds. Amsterdam, The Netherlands: Springer-Verlag, Oct. 2002, vol. 2489, pp. Carried out according to the techniques described in 1-11, the entirety of this article is hereby incorporated by reference.

4.6 Automatic Evaluation of Tree Quality Tree quality is evaluated based on recognition of the following key parts of the skeleton:
4.6.1 Lower right leaf (RLL) and right middle leaf (RML),
4.6.2 Upper right leaf (RUL)
4.6.3 Upper left leaf (LUL)
4.6.4 Lower left leaf (LLL)
The branch number and branch length features are calculated separately for each area and compared to a statistical model or acceptable anatomy template to assess tree quality.

4.7 Extraction of the external surface of the bronchial tube In order to extract the airway surface from the volumetric CT data, a modification of a widely known method called the “marching cube” is used.

5). Planning the path to the peripheral target This process plans the path from the tracheal entrance to the target area.

  Since the CT resolution limits the final quality of the automatically generated bronchial tree described at 1, the user can perform fine tuning of the path.

6). Target marking The planning software is used to plan a bronchoscopic procedure to navigate to a lesion (target) in the human lung.

  The target center and target dimensions are marked manually using the planning software.

7). Semi-automatic generation of paths There are both automatically generated bronchial trees and targets at this point. However, the target can be located outside the tree. This occurs for several reasons including:

  -The target may be inside the tissue.

  -Some small bronchi may disappear from the automatically generated tree due to CT resolution limitations.

  Thus, a gap is created and manually completed. Using both the interactive display of the bronchial tree and the CT cross section, the user manually selects a point on the bronchial tree that is connected to the center of the target. This is referred to as an “exit point”.

  A route from the trachea to the “exit point” is automatically generated. In addition, the original tree is extended by a straight branch connecting the “exit point” and the center of the target.

8). Fine-tuning the path CT cross-sections allow the user to define an intermediate waypoint through an intuitive graphic user interface and optionally divide automatically created straight branches into segments Can be defined.

9. Path Guidance Path Guidance is designed to hold a tool that can be placed inside the planned path (shown in green). In this approach, the path is approximated by an automatically generated 3D poly-line. The polyline segment is connected to the vertex. Each vertex is defined as an intervening target in our system. During navigation, when an intervening target is reached during navigation, it disappears and the next intervening target appears and becomes the current target. A mathematical vector is calculated that connects the actual position of the installable tool to the successor intervening target. This mathematical vector is translated into a tool operation that can be installed via the following instruction set:
9.1 Push forward backward 9.2 Set specific rotation angle 9.3 Apply bending ON \ OFF.
Additional methods are contemplated that can improve the accuracy of the BT produced by the above method. Initially, it can be used to regenerate airway data in which arterial vessels are detected and missing from the CT. Since the arterial blood vessels from the heart to the lungs terminate in the alveoli, a deduction can be made regarding the location of the bronchioles leading to the alveoli. Second, an anatomical atlas created from data derived from multiple lung models can be used to evaluate and complete the generated BT geometry. All lungs are unique, but each lung has a common characteristic drawn in the anatomical atlas. This information can be used to deduct and fill in missing BT geometry data.

  Accuracy can also be improved by utilizing multiple sensors. For example, obtaining position and orientation measurement data from an electromagnetic system can be performed using multiple external and / or multiple internal sensors. They can be located on a long working channel (EWC), a placeable guide (LG), a bronchoscope, or attached inside the lungs.

  Regardless of the number of sensors used, the position and orientation measurement data obtained from the electromagnetic system can be used to complete any branches that are missing from the BT due to CT resolution limitations.

  It is also contemplated that flexibility may be added to the generated BT structure by utilizing multiple CT data sets, where each data set represents a different point in the patient's respiratory cycle. For example, three CT scans may be taken, one with a peak inspiration point during a normal breathing cycle, one with a peak exhalation point during a normal breathing cycle, and one between the two. The external sensor portion can optionally be noted to record the chest during these various “snapshots” taken with CT. Note the difference in the location of the bronchial feature at each of the three locations, providing information about each of the feature movement paths during the respiratory cycle. The path of motion can be estimated by connecting the three recorded points. Once the flexible BT is generated, the patient's external position sensor determines the corresponding position of the various bronchial features along the path of their respective movements to detect the patient's respiratory cycle and Can be used to

  This simulated flexibility can be calculated and used individually for each patient, or other patients if it is desired to minimize the cost and radiation exposure of multiple CT scans. Can be used as a model for Several models can be recorded and stored in a file for later adaptation to the patient as a function of anatomical location, patient size, sex age, respiratory cycle phase, and the like.

(Adaptive navigation method)
The present invention also provides a unique method of continuously and adaptively matching a patient with a BT skeleton that is automatically generated during the procedure. In general, this method records the continuous placement of the LG as it is advanced through the airway. Since LG is known to move through the airway, the BT skeleton is continuously matched so that LG appears in the airway. Thus, the accuracy of navigation improves without going down as the LG is advanced.

More specifically, this method is outlined as follows:
1. General Considerations An adaptive skeleton navigation method detects the current position of a placeable guide (“LG”) introduced through a patient's bronchial airway on a map of the bronchial tree acquired using a CT scan. Developed for. This is achieved by a constant and adaptive correlation between the two data sets, the bronchial airway tree map and the sensor data history. The above correlation is performed through two steps.

  1) Adaptive skeleton based registration.

  2) Adaptive skeleton-based navigation.

Note that these steps described in detail below may be performed iteratively.
2. Adaptive Skeleton-Based Registration Algorithm This section contains a description of the proposed algorithm for adaptive skeleton-based registration. Registration is generally a way to calculate the transformation between two different coordinate systems. Here, the purpose is to register the bronchial tree (BT) skeleton into a guide (LG) path where it can be placed.

2.1 Requirements:
2.1.1 Registration accuracy improves as the deployable guide approaches the lower level lumen network (eg, bronchial tree) and peripheral targets.

  2.1.2 The registration is updated continuously and adaptively depending on the position of the LG in the bronchial tree. The LG path is a history of the position of the LG as it is manipulated through the bronchial tree.

2.2 Technical issues:
2.2.1 Geometric 3D / 3D point (or other object) pairs from the BT skeleton and LG path are the basis of registration (eg, pair 1-1 ′ and See 2-2 ').

  2.2.2 Registration is continuous and adaptive. “Continuous” means that the registration is continuously (repetitively) recalculated as new LG path points are acquired. “Adaptive” means that different pairs of points, weights, and registration methods are used as the LG progresses toward the target.

  2.2.3 Registration consists of two main stages: an overall fixed registration followed by a local deformable registration. Deformable registration is performed only at the lower level bronchial tree and near the peripheral target. The idea is to start a fixed registration when the bronchi is wide, and to a constrained localized deformable registration when the probe diameter approaches the bronchial diameter and the bronchus becomes flexible. It is to switch.

  2.2.4 The entire fixed registration is performed by a highly weighted iterative closest point (WICP) method with outlier removal. Pairing is performed by using a weighted function of the distance between the positions of the paired objects and the orientation difference. The optimization function is a weighted sum of paired point distances. The parameters to be determined are the weight function and the number of points to be used.

  2.2.5 Local non-fixed registration is performed using a confined elastic registration method, the paired points are connected to the spring and the optimization function is the spring potential energy .

  2.2.6 The LG path is not monotonous, so it should be carefully sampled and windowed so that the path data is of good quality.

  2.2.7 Accuracy or registration improves when user-defined landmarks are obtained with LG.

3. Registration algorithm 3.1 Input: BT path from CT scan, first registration guess, LG location (stream)
3.2 Output: Fixed Registration (6 parameters) + Local deformation map 3.3 Algorithm Method-FIG. 1 shows a line showing the BT skeleton 110 and a second line 120 showing the path of LG 115 through the bronchus 100 A schematic of an actual bronchus 100 having The LG 115 path 120 is used to register the BT skeleton 110 to the actual bronchus 100 so that the BT skeleton seen by the physician is an accurate representation of the actual bronchial position. The LG path 120 includes a plurality of actual LG locations 130. A corresponding projected point 140 is shown on the BT skeleton 110. The difference between the actual point 130 and the projected point 140 is represented by a line (eg, d1, d2) between the skeleton 110 and the path 120. Subsequent reference to FIG. 1 illustrates the registration algorithm:
3.3.1 Perform the first registration with the first registration guess. Apply deformation to BT. Use the registration results obtained from the initial registration stage.

3.3.2 While on the way to the target:
3.3.2.1 Obtain a new stream of LG positions from the sensor.

  3.3.2.2 Perform cleaning, scattering (scatter) and classification (weighting) on the LG position stream.

  3.3.2.3 Perform LG position stream selection according to optimization decisions and registration history.

  3.3.2.4 Project the selected LG location segment / point onto the BT skeleton to obtain the paired segment / point. Projection is performed by optimizing the following criteria:

  3.3.4.2.4.1 Minimum distance to local bronchial diameter.

  3.3.4.2.4.2 Minimum orientation difference.

  3.3.4.2.3 Matched branch points on the path (p1, p2, etc.).

  3.3.2.5 Overall Fixed Registration: Weight the paired points to get a new fixed registration with WICP. Apply the new calculated deformation to LG.

  3.3.2.6 Local deformable registration: If appropriate (only when long working channel-EWC and bronchial diameter are close to each other), perform deformable registration on the selected window . The transformation is applied to the BT local branch. The usefulness of this response is determined empirically.

3.3.3 Enabling registration not to be bad as a result of noise sensor data:
3.3.3.1 Perform LG history classification by breathing averaging or specific aspects. Define the maximum deviation from the initial registration.

  3.3.3.2 Make sure that the LG history is mostly inside the bronchi.

4). Adaptive Skeleton-Based Navigation 4.1 Basic Concepts Navigation implementations will be similar to map navigation. Similar tasks have been implemented in GIS (Geographic Information Systems) such as PDA (Personal Digital Assistant) systems for blind people. The data has proven that topology is important for higher navigation accuracy. However, due to the flexibility and movement of the bronchial tree, the problem is substantially different from the above.

4.2 Required Input Information 4.2.1 Current Sensor Position and Orientation Data 4.2.2 History of Sensor Position and Orientation Data 4.2.3 Registration (Matrix) History 2.4 Sources of Navigation Uncertainty FIG. 2 is a table showing the sources of error uncertainty in importance ranking, with values having a higher contribution to the final error at the top of the table. An error prediction model based on this table will be developed to predict navigation uncertainty. The assumed prediction model is a sphere that encloses the calculated position, and the radius of the sphere includes position uncertainty. The radius of this sphere is a function of time and position.

(In use)
3-13 show several steps of the above method. First, the path generation process is described. FIG. 3 is a chart showing the route generation steps 10 to 18. 4-13 are user interface displays that guide the user through these steps during the planning phase prior to the procedure.

  The user engaged in the first step 10 adding a target is shown in FIG. The upper left quadrant is a CT sectional view reconstructed from the viewpoint of looking at the head from the patient's foot. The lower left quadrant is a CT cross section reconstructed from the front of the patient so that the plane is parallel to the platform on which the patient lies. The upper right quadrant is a cross-sectional view reconstructed from CT directed to the patient's side. The crosshairs for aiming mark the targeted spot, which is the intersection of all cross-sectional views reconstructed from CT. Each of the CT cross sections shows a small camera projection representing a virtual bronchoscope in the CT volume. The lower right quadrant is a virtual bronchoscopic image of a targeted spot from inside the BT, as seen from the camera. It is worth noting that the display shown in the figure is just an example. If the user feels the need for different images, the system allows images from any angle to be displayed in various quadrants. Thus, the system is fully configurable and can be customized to user preferences.

  In FIG. 5, the user has selected the target and the message “Aiming crosshair positioning” appears in the user message box in the upper right corner of the user interface. As can be seen in FIG. 6, selecting the very bottom button 20 marks the target. The user may enter the target name in box 22 and thus begin the next step of FIG.

  FIG. 7 illustrates that the user has the option to display the BT skeleton in the upper right quadrant rather than the side image.

  In FIG. 8, the selected target is being measured. In FIG. 9, the name is given to the target.

  In FIG. 10, the route creation step 14 begins. The route button 24 is pressed and a name is given to the route.

  FIG. 11 shows the exit point marking step 16. The exit point is marked on the map and a straight line is drawn from the exit point to the target. Once the destination (exit point) has been identified, the route can be determined.

  Then, as seen in FIGS. 12 and 13, the step 18 of adding waypoints is complete. An intermediate point can be added to help the user follow the path by marking a corner on the way to the exit point.

  FIG. 14 shows the user interface during the procedure being performed on the patient. In the upper right quadrant, a three-dimensional BT skeleton is shown with the LG indicator 30 visible. The above AN algorithm ensures that the BT skeleton remains registered to the patient.

  Although the present invention has been described with respect to particular embodiments and applications, those skilled in the art will be able to make additional embodiments and modifications in light of this teaching without departing from the spirit or scope of the claimed invention. Can be generated. Accordingly, it should be understood that the figures and descriptions herein are provided as examples to facilitate understanding of the invention and should not be construed as limiting the scope of the invention. It is. Thus, it should be noted that any additional documents referenced in the attached documents are hereby incorporated by reference in their entirety.

Claims (14)

A method of generating an intraluminal pathway to a target inside the body, the method comprising:
Imaging the patient to obtain multiple scans;
Creating a 3D image volume from the scan including a plurality of voxels;
Selecting a target in the 3D image volume;
Detecting anatomical features related to the network;
Segmenting voxels representing the space inside the lumen of the network;
Generating a computerized three-dimensional model of the lumen network using the segmented voxels;
Marking a route to the target on a three-dimensional model of the lumen network.   The method of claim 1, wherein detecting an anatomical feature associated with the network includes detecting a trachea.   The method of claim 2, wherein detecting the trachea includes looking for a tubular object in the lumen network having a density of air in an upper region of the network.   The method of claim 1, wherein detecting an anatomical feature associated with the network comprises detecting an aorta.   The segmenting voxels that represent the space inside the lumen of the network comprises marking voxels that represent material contained within the lumen of the network. The method described in 1.   The method of claim 5, wherein segmenting voxels representing material inside the lumen of the network further comprises avoiding bubbles caused by imaging noise and artifacts. Segmenting voxels representing the space inside the lumen of the network comprises
a) defining a threshold;
b) defining a seed point at the center point inside the anatomical feature related to the network;
c) marking all relevant voxels starting from the seed point according to the threshold;
d) recording the total number of marked voxels;
e) increasing the threshold;
f) repeating steps c) and d) thereby starting the next iteration;
g) comparing the number of marked voxels with the number determined in the previous iteration;
h) Set the limit for the increase determined in g) where the increase is due to a leak, and the threshold increased in e) is reset from the previous iteration for that limit and selected accordingly Constructing a determined threshold;
i) segmenting the marked voxels with the selected threshold;
j) adding the segmented voxel to the seed point, thereby growing a path, and resetting the seed point to the center of the distal end of the path;
2. The method of claim 1, comprising: k) repeating steps c) to j) until the path is complete. A method of registering a computer model of a lumen network with a patient's actual lumen network while navigating the lumen network with a probe, the method comprising:
Collecting data on the actual position of the probe;
Processing consecutive data points to construct a probe path;
Repetitively registering the path of the probe in the form of a path on a computer model of a lumen network while tracking the position of the probe;
Repetitively adjusting the position of the computer model such that the path of the probe is within the shape of the path.   The method of claim 8, wherein collecting data relating to an actual position of the probe includes receiving position data from a tip of the probe.   The method of claim 8, wherein processing successive data points to construct a probe path includes cleaning.   Iteratively registering the probe path in the form of a path on a computer model of the lumen network while tracking the position of the probe results in the path of the probe being tracked while tracking the position of the probe. 9. The method of claim 8, comprising fixed registration in the form of a path on the computer model.   Iteratively registering the probe path in the form of a path on a computer model of the lumen network while tracking the position of the probe results in the path of the probe being tracked while tracking the position of the probe. 9. The method of claim 8 comprising deformably registering in the form of a path on a computer model of the computer. Iteratively registering the path of the probe into the form of a path on a computer model of a lumen network while tracking the position of the probe
Registering the path of the probe fixedly in the form of a path on a computer model of the lumen network while tracking the position of the probe while the diameter of the lumen containing the probe is large relative to the probe; ,
While the diameter of the lumen containing the probe is not large relative to the probe, the probe path is deformably registered into a path on the computer model of the lumen network while tracking the position of the probe. The method according to claim 8, comprising: Iteratively registering the path of the probe into the form of a path on a computer model of a lumen network while tracking the position of the probe
Acquiring a stream of probe positions;
Contain the scattering state of the probe position, weight the probe position,
Selecting the position of the probe for projection;
9. The method of claim 8, comprising projecting the position of the selected probe onto a computer model.