SYSTEM AND METHOD FOR PASSIVE MEDICAL DEVICE NAVIGATION UNDER REAL-TIME MRI GUIDANCE

• US 20120089008A1
• Filed: 06/01/2011
• Published: 04/12/2012
• Est. Priority Date: 06/09/2010
• Status: Active Grant

• Current Assignee: Siemens Healthineers AG (Siemens AG)
• Original Assignee: Siemens Corp. (Siemens AG)
• Inventors: Strehl, Wilhelm, Gilson, Wesley D., Rothgang, Eva
• Granted Patent: US 8,526,691 B2
• Time in Patent Office: Days
• US Class Current: 600/411
• CPC Class Codes: A61B 2034/2065 Tracking using image or pattern recognition
• CPC Class Codes: A61B 2090/374 NMR or MRI
• CPC Class Codes: A61B 34/20 Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
• CPC Class Codes: G01R 33/286 involving passive visualization of interventional instruments, i.e. making the instrument visible as part of the normal MR process

Abstract

A method for tracking a needle along a path under real-time magnetic resonance imaging (MRI) guidance includes extracting a candidate needle artifact region from a plurality of incoming MR images of a needle inside biological tissue, determining an initial needle direction vector from the candidate needle artifact region in each of the plurality of incoming needle images, determining an initial needle base point from an intersection of the needle direction vector with a first incoming needle image, and an initial needle tip point from an intersection of the needle direction vector with a last incoming needle image, and tracking said needle by minimizing a cost function of the initial base point and initial direction vector to calculate an updated base point and updated direction vector. The updated base point and direction vector are used to visualize the advance of the needle along the path through the biological tissue.

Claims

1. A method for tracking a needle along a path under real-time magnetic resonance imaging (MRI) guidance, comprising the steps of:

extracting a candidate needle artifact region from a plurality of incoming MR images of a needle inside biological tissue;

determining an initial needle direction vector from the candidate needle artifact region in each of the plurality of incoming needle images;

determining an initial needle base point from an intersection of the needle direction vector with a first incoming needle image, and an initial needle tip point from an intersection of the needle direction vector with a last incoming needle image; and

tracking said needle by minimizing a cost function of the initial base point and initial direction vector to calculate an updated base point and updated direction vector, wherein the updated base point and direction vector are used to visualize the advance of the needle along the path through the biological tissue.

2. The method of claim 1, wherein extracting candidate needle artifact regions comprises:

receiving a current MR needle image defined as an MR image containing the needle artifact;

extracting a plurality of pixel regions from the current MR needle image having a lower image intensity than surrounding pixels;

calculating a plurality of features for each pixel region; and

selecting one or more pixel regions, based on the features of each region, that are likely to represent a needle artifact, as candidate needle artifact regions,

wherein, if more than one pixel region is selected, choosing a region closest to the path as the candidate needle artifact region.

3. The method of claim 2, wherein the plurality of pixel regions are extracted from the current MR needle image by subtracting the current MR needle image from a background model, extracting those pixels whose intensity difference with respect to corresponding pixels in the background model is greater than a predetermined threshold, and using connected-component analysis to form pixel regions from the extracted pixels.

4. The method of claim 3, wherein a pixel i is extracted if the intensity difference E_t(x_i)−

I_t(x_i)≧

d_threshold, wherein E_t is the background model, I_t is the current image, and x_i is the position of pixel i.

5. The method of claim 3, further comprising, if no region is selected as a candidate needle artifact region, updating the background model with the current image.

6. The method of claim 2, wherein features calculated for the extracted pixel regions include an area, a perimeter, a circularity, a centroid, and a distance from the centroid to the planned path, wherein regions whose centroid is less than a predetermined distance from the planned path, whose circularity is less than a predetermined maximum, and whose area is within a predetermined range, are selected as candidate needle artifact regions.

7. The method of claim 1, wherein said cost function of the base point and direction vector is given by

8. The method of claim 7, wherein minimizing said cost function comprises acquiring at least three images with respect to a current needle position, wherein at least two images are parallel to the needle and perpendicular to each other, and the third image is perpendicular to longitudinal axis of the needle, and minimizing said cost function in said images to obtain a new needle position.

9. The method of claim 7, wherein minimizing said cost function comprises updating a parameter vector p=(b, d) according to

10. The method of claim 7, wherein minimizing said cost function comprises updating a parameter vector p=(b, d) according to p_i+1=p_i−

γ

_GD∇

J(p_i), where γ

_GD denotes a step size.

11. A method for tracking a needle along a path under real-time magnetic resonance imaging (MRI) guidance, comprising the steps of:

determining an initial base point and an initial direction vector of a needle during an MRI guided intervention by analyzing a needle artifact region in each of a plurality of incoming MR images; and

tracking said needle during said intervention by minimizing a cost function of the initial base point and initial direction vector to calculate an updated base point and updated direction vector, wherein said cost function of the base point and direction vector is given by

12. The method of claim 11, wherein determining an initial base point and an initial direction vector of a needle comprises:

receiving a current MR needle image;

extracting a plurality of pixel regions from the current MR needle image having a lower image intensity than surrounding pixels;

calculating a plurality of features for each pixel region; and

selecting one or more pixel regions, based on the features of each region, that are likely to represent a needle artifact, as needle artifact regions,

wherein, if more than one pixel region is selected, choosing a region closest to the path as the needle artifact region.

13. The method of claim 12, wherein determining a base point and a direction vector of a needle comprises:

determining an initial needle direction vector from the needle artifact region in each of the plurality of incoming MR images; and

determining an initial needle base point from an intersection of the needle direction vector with a first incoming MR image, and an initial needle tip point from an intersection of the needle direction vector with a last incoming MR image.

14. A program storage device readable by a computer, tangibly embodying a program of instructions executable by the computer to perform the method steps for tracking a needle along a path under real-time magnetic resonance imaging (MRI) guidance, the method comprising the steps of:

extracting a candidate needle artifact region from a plurality of incoming MR images of a needle inside biological tissue;

determining an initial needle direction vector from the candidate needle artifact region in each of the plurality of incoming needle images;

determining an initial needle base point from an intersection of the needle direction vector with a first incoming needle image, and an initial needle tip point from an intersection of the needle direction vector with a last incoming needle image; and

tracking said needle by minimizing a cost function of the initial base point and initial direction vector to calculate an updated base point and updated direction vector, wherein the updated base point and direction vector are used to visualize the advance of the needle along the path through the biological tissue.

15. The computer readable program storage device of claim 14, wherein extracting candidate needle artifact regions comprises:

receiving a current MR needle image defined as an MR image containing the needle artifact;

extracting a plurality of pixel regions from the current MR needle image having a lower image intensity than surrounding pixels;

calculating a plurality of features for each pixel region; and

selecting one or more pixel regions, based on the features of each region, that are likely to represent a needle artifact, as candidate needle artifact regions,

wherein, if more than one pixel region is selected, choosing a region closest to the path as the candidate needle artifact region.

16. The computer readable program storage device of claim 15, wherein the plurality of pixel regions are extracted from the current MR needle image by subtracting the current MR needle image from a background model, extracting those pixels whose intensity difference with respect to corresponding pixels in the background model is greater than a predetermined threshold, and using connected-component analysis to form pixel regions from the extracted pixels.

17. The computer readable program storage device of claim 16, wherein a pixel i is extracted if the intensity difference E_t(x_i)−

I_t(x_i)≧

d_threshold, wherein E_t is the background model, I_t is the current image, and x_i is the position of pixel i.

18. The computer readable program storage device of claim 16, the method further comprising, if no region is selected as a candidate needle artifact region, updating the background model with the current image.

19. The computer readable program storage device of claim 15, wherein features calculated for the extracted pixel regions include an area, a perimeter, a circularity, a centroid, and a distance from the centroid to the planned path, wherein regions whose centroid is less than a predetermined distance from the planned path, whose circularity is less than a predetermined maximum, and whose area is within a predetermined range, are selected as candidate needle artifact regions.

20. The computer readable program storage device of claim 14, wherein said cost function of the base point and direction vector is given by

21. The computer readable program storage device of claim 20, wherein minimizing said cost function comprises acquiring at least three images with respect to a current needle position, wherein at least two images are parallel to the needle and perpendicular to each other, and the third image is perpendicular to longitudinal axis of the needle, and minimizing said cost function in said images to obtain a new needle position.

22. The computer readable program storage device of claim 20, wherein minimizing said cost function comprises updating a parameter vector p=(b, d) according to

23. The computer readable program storage device of claim 20, wherein minimizing said cost function comprises updating a parameter vector p=(b, d) according to p_i+1=p_i−

γ

_GD∇

J (p_i) where γ

_GD denotes a step size.