System and method for detecting and tracking a guidewire in a fluoroscopic image sequence

1. A method for populating a database with a set of image sequences of an object the database being used to detect localization of a guidewire in the object, the method comprising the steps of:

receiving a set of images of anatomical structures, each image being annotated to show a guidewire, catheter, wire tip and stent;

for each given image:

using a Probabilistic Boosting Tree (PBT) to detect short line segments of constant length in the image;

constructing two segment curves from the short line segments;

using a discriminative joint shape and appearance model to classify each two segment curve;

constructing a shape of an n-segment curve by concatenating all the two segment curves;

identifying a guidewire curve model that comprises a start point, end point and the n-segment curve; and

storing the guidewire curve model in the database.

2. The method of claim 1 wherein the guidewire is represented as a one dimensional curve in a three dimensional space.

3. The method of claim 1 wherein a positive sample comprises a short line segment of constant length.

4. The method of claim 3 wherein the short line segment comprises a segment center location and a segment orientation.

5. The method of claim 4 wherein the segment orientation is discretized into a plurality of discrete image rotation values.

6. The method of claim 5 wherein the image rotation values are represented by Haar features.

7. The method of claim 1 wherein a number of two segment curves detected is controlled by the number of positive samples detected by the segment detector.

8. The method of claim 1 wherein a probability is computed for each two segment curve to determine if the two segment curve is part of the guidewire.

9. The method of claim 1 wherein a Principal Component Analysis model is used to model the shape of each two segment curve.

10. The method of claim 1 wherein the object is an anatomical structure.

11. The method of claim 10 wherein the anatomical structure is a heart.

12. The method of claim 1 wherein the images are fluoroscopic images.

13. A system for detecting localization of a guidewire in an anatomical structure in an image sequence, the system comprising:

a database that stores a set of image sequences of anatomical structures, each image being annotated to annotated to show a guidewire, catheter, wire tip and stent;

a processor that receives an input image sequence of an anatomical structure, the processor performing the following steps on each image in the image sequence:

using a Probabilistic Boosting Tree (PBT) to detect short line segments of constant length in the image;

constructing two segment curves from the short line segments;

using a discriminative joint shape and appearance model to classify each two segment curve;

constructing a shape of an n-segment curve by concatenating all the two segment curves;

identifying a guidewire curve model that comprises a start point, end point and the n-segment curve;

constructing the guidewire using the guidewire curve model; and

a display to display the guidewire in the anatomical structure.

14. The system of claim 13 wherein the guidewire is represented as a one dimensional curve in a three dimensional space.

15. The system of claim 13 wherein a positive sample comprises a short line segment of constant length.

16. The system of claim 15 wherein the short line segment comprises a segment center location and a segment orientation.

17. The system of claim 16 wherein the segment orientation is discretized into a plurality of discrete image rotation values.

18. The system of claim 17 wherein the image rotation values are represented by Haar features.

19. The system of claim 13 wherein a number of two segment curves detected is controlled by the number of positive samples detected by the segment detector.

20. The system of claim 13 wherein a probability is computed for each two segment curve to determine if the two segment curve is part of the guidewire.

21. The system of claim 13 wherein a Principal Component Analysis model is used to model the shape of each two segment curve.

22. The system of claim 13 wherein the anatomical structure is a heart.

23. The system of claim 13 wherein the images are fluoroscopic images.