Device and method for automatically detecting a surgical tool on an image provided by a medical imaging system

1. A method of detecting a surgical tool in, a first image comprising the steps of:

storing shape parameters of the surgical tool;

determining a second image containing distinct pixel regions from the first image by a first processing comprising segmenting the first image;

determining first windows, each first window surrounding one of the regions;

selecting at least one second window from among the first windows based on the shape parameters;

determining a third image from the first image by applying to the first image a second processing implementing the Frangi algorithm; and

determining lateral edges of the surgical tool in the portion of the third image contained in said second window.

2. The method of claim 1 , wherein the surgical tool has an elongated shape, the lateral edges of the surgical tool in the third image corresponding to straight lines.

3. The method of claim 2 , wherein candidate straight lines classified according to a ranking order are determined by applying a Hough transform in the portion of the third image contained in the second window and wherein the lateral edges are determined, among the candidate straight lines based on the shape parameters.

4. The method of claim 3 , wherein the second window is rectangular and divided into a first sub-window and a second sub-window, wherein a first lateral edge corresponds to the candidate straight line having the best rank in the first sub-window and having an inclination with respect to the major axis of symmetry of the second window smaller than a first threshold, and wherein a second lateral edge corresponds to the candidate straight line having the best rank in the second sub-window, having an inclination with respect to the major axis of symmetry of the second window smaller than the first threshold, and having its distance to the first lateral edge in the range from a second threshold to a third threshold which depend on the width of the second window.

5. The method of claim 1 , wherein the first processing comprises binarizing the first image and determining the second image based on the determination of a distance map of the binarized image.

6. The method of claim 1 , wherein the second processing comprises applying the Frangi algorithm to the first image with a search parameter, representative of the number of pixels of the apparent diameter of tubular structures searched for by the Frangi algorithm, smaller than 10 pixels.

7. The method of claim 1 , wherein the first image is an image in grey levels obtained from component ‘a’ or ‘b’ of a fourth image acquired by a camera and coded in the CIE Lab color space.

8. The method of claim 1 , wherein the first windows are rectangular and wherein the second window is selected by only retaining the first windows having a length greater than half their width.

9. The method of claim 1 , wherein the first image is obtained from a fourth image acquired by a camera, the method further comprising the steps of:

determining projection planes containing the lateral edges and a projection center of the camera to provide two projection planes; and

determining the position of a central axis of the surgical tool based on the intersection of the two projection planes, on the shape parameters, and on a point belonging to the central axis.

10. The method of claim 9 , further comprising the steps of:

determining a projection of the central axis in the third image;

determining the pixels of the projection of the central axis belonging to the surgical tool; and

determining the pixel of the projection of the central axis corresponding to the end of the tool.

11. The method of claim 10 , comprising the steps of:

assigning to each pixel of the third image of the projection of the central axis a color class from among a first or a second color class;

selecting a first, group of adjacent pixels containing the largest number of adjacent pixels of the first color class and a second group of adjacent pixels containing the largest number of adjacent pixels of the second color class; and

selecting, from among the first or the second group of pixels, the pixel group belonging, to the surgical tool.

12. A device for selecting a surgical tool in a first image (I t ), comprising:

means for storing shape parameters of the surgical tool;

means for determining a second image containing distinct pixel regions based on the first image by a first processing comprising segmenting the first image;

means for determining first windows, each window surrounding one of the regions;

means for selecting at least one second window from among the first windows based on the shape parameters;

means for determining a third image based on the first image by a second processing enhancing the contours of the surgical tool in the first image; and

means for determining lateral edges of the surgical tool on the portion of the third image contained in said second window.