Method and system for obtaining a sequence of x-ray images using a reduced dose of ionizing radiation

1. A method for obtaining a sequence of x-ray images, comprising:

detecting an object of interest in a patient in a first x-ray image;

determining an area of interest based on a predicted motion of the object of interest;

acquiring a second x-ray image based on the area of interest determined in the first x-ray image using spatial x-ray modification to control an x-ray to pass through a portion of the patient corresponding to the area of interest, the second x-ray image having a reduced imaging size of the patient relative to the first x-ray image;

determining whether the object of interest was successfully detected in the second x-ray image; and

in response to determining that the object of interest was not successfully detected in the second x-ray image:

generating an x-ray scatter image of a region beyond the area of interest determined in the first x-ray image.

2. The method of claim 1 , wherein detecting an object of interest in a first x-ray image comprises:

automatically detecting the object of interest in the first x-ray image using a probabilistic boosting tree (PBT) trained on annotated training data.

3. The method of claim 2 , wherein detecting an object of interest in a first x-ray image further comprises:

determining at least one low probability region of the first x-ray image based on information related to a guidance path of the object of interest; and

removing the determined at least one low probability region from a search space of the PBT prior to automatically detecting the object of interest in the first x-ray image using the PBT.

4. The method of claim 2 , wherein detecting an object of interest in a first x-ray image further comprises:

applying a steerable filter to the first x-ray image to identify at least one high contrast region of the first x-ray image; and

removing the identified at least one high contrast region from a search space of the PBT prior to automatically detecting the object of interest in the first x-ray image using the PBT.

5. The method of claim 1 , wherein determining an area of interest based on a predicted motion of the object of interest comprises:

predicting a next location of the object of interest by predicting a motion of the object of interest; and

defining the area of interest as a region surrounding the predicted next location of the object of interest.

6. The method of claim 5 , wherein predicting a next location of the object of interest by predicting a motion of the object of interest comprises:

predicting the motion of the object of interest using one of: an Extended Kalman Filter model, a Particle Filter model, and a Learnt Motion model.

7. The method of claim 6 , wherein the one of the Extended Kalman Filter model, the Particle Filter model, and the Learnt Motion model incorporates prior information regarding a planned path of the object of interest.

8. The method of claim 1 , wherein acquiring a second x-ray image of the area of interest using spatial x-ray modification to control an x-ray to pass through a portion of a patient corresponding to the area of interest comprises:

changing at least one collimator of an imaging device to alter at least one of a shape and a position of an x-ray lightbeam, such that the x-ray lightbeam only passes through the portion of the patient corresponding to the area of interest.

9. The method of claim 1 , wherein acquiring a second x-ray image of the area of interest comprises:

utilizing at least one semitransparent collimator of the image device to constrain the x-ray lightbeam to only pass through the portion of the patient corresponding to the area of interest.

10. The method of claim 1 , wherein acquiring a second x-ray image of the area of interest comprises:

changing at least one of an angulation of a C-arm of the imaging device and a rotation of a detector of the imaging device.

11. The method of claim 1 , wherein acquiring a second x-ray image of the area of interest comprises:

changing a position of a table of the imaging device.

12. The method of claim 1 , further comprising:

detecting the object of interest in the second x-ray image;

determining a second area of interest based on a second predicted motion of the object of interest detected in the second x-ray image; and

acquiring a third x-ray image of the second area of interest using spatial x-ray modification to control an x-ray to pass through a portion of the patient corresponding to the second area of interest.

13. The method of claim 1 , further comprising:

in response to determining that the object of interest was not successfully detected in the second x-ray image:

detecting the object of interest in the x-ray scatter image.

14. The method of claim 13 , wherein detecting the object of interest in the x-ray scatter image comprises:

detecting the object of interest in the x-ray scatter image using a probabilistic boosting tree (PBT) trained on annotated training data comprising x-ray scatter image data.

15. A non-transitory computer readable medium storing computer program instructions for obtaining a sequence of x-ray images, which, when the instructions are executed on a processor, cause the processor to perform operations comprising:

detecting an object of interest in a patient in a first x-ray image;

determining an area of interest based on a predicted motion of the object of interest;

acquiring a second x-ray image based on the area of interest determined in the first x-ray image using spatial x-ray modification to control an x-ray to pass through a portion of the patient corresponding to the area of interest, the second x-ray image having a reduced imaging size of the patient relative to the first x-ray image;

determining whether the object of interest was successfully detected in the second x-ray image; and

in response to determining that the object of interest was not successfully detected in the second x-ray image:

generating an x-ray scatter image of a region beyond the area of interest determined in the first x-ray image.

16. The non-transitory computer readable medium of claim 15 , wherein detecting an object of interest in a first x-ray image comprises:

automatically detecting the object of interest in the first x-ray image using a probabilistic boosting tree (PBT) trained on annotated training data.

17. The non-transitory computer readable medium of claim 16 , wherein detecting an object of interest in a first x-ray image further comprises:

determining at least one low probability region of the first x-ray image based on information related to a guidance path of the object of interest; and

removing the determined at least one low probability region from a search space of the PBT prior to automatically detecting the object of interest in the first x-ray image using the PBT.

18. The non-transitory computer readable medium of claim 16 , wherein detecting an object of interest in a first x-ray image further comprises:

applying a steerable filter to the first x-ray image to identify at least one high contrast region of the first x-ray image; and

removing the identified at least one high contrast region from a search space of the PBT prior to automatically detecting the object of interest in the first x-ray image using the PBT.

19. The non-transitory computer readable medium of claim 15 , wherein determining an area of interest based on a predicted motion of the object of interest comprises:

predicting a next location of the object of interest by predicting a motion of the object of interest; and

defining the area of interest as a region surrounding the predicted next location of the object of interest.

20. The non-transitory computer readable medium of claim 19 , wherein predicting a next location of the object of interest by predicting a motion of the object of interest comprises:

predicting the motion of the object of interest using one of: an Extended Kalman Filter model, a Particle Filter model, and a Learnt Motion model.

21. The non-transitory computer readable medium of claim 20 , wherein the one of the Extended Kalman Filter model, the Particle Filter model, and the Learnt Motion model incorporates prior information regarding a planned path of the object of interest.

22. The non-transitory computer readable medium of claim 15 , wherein acquiring a second x-ray image of the area of interest using spatial x-ray modification to control an x-ray to pass through a portion of a patient corresponding to the area of interest comprises:

changing at least one collimator of an imaging device to alter at least one of a shape and a position of an x-ray lightbeam, such that the x-ray lightbeam only passes through the portion of the patient corresponding to the area of interest.

23. The non-transitory computer readable medium of claim 15 , wherein acquiring a second x-ray image of the area of interest comprises:

utilizing at least one semitransparent collimator of the image device to constrain the x-ray lightbeam to only pass through the portion of the patient corresponding to the area of interest.

24. The non-transitory computer readable medium of claim 15 , wherein acquiring a second x-ray image of the area of interest comprises:

changing at least one of an angulation of a C-arm of the imaging device of the imaging device and a rotation of a detector of the imaging device.

25. The non-transitory computer readable medium of claim 15 , wherein acquiring a second x-ray image of the area of interest comprises:

changing a position of a table of the imaging device.

26. The non-transitory computer readable medium of claim 15 , further comprising:

detecting the object of interest in the second x-ray image;

determining a second area of interest based on a second predicted motion of the object of interest detected in the second x-ray image; and

acquiring a third x-ray image of the second area of interest using spatial x-ray modification to control an x-ray to pass through a portion of the patient corresponding to the second area of interest.

27. The non-transitory computer readable medium of claim 15 , further comprising:

in response to determining that the object of interest was not successfully detected in the second x-ray image:

detecting the object of interest in the x-ray scatter image.

28. The non-transitory computer readable medium of claim 27 , wherein detecting the object of interest in the x-ray scatter image comprises:

detecting the object of interest in the x-ray scatter image using a probabilistic boosting tree (PBT) trained on annotated training data comprising x-ray scatter image data.

29. A system for obtaining a sequence of x-ray images, comprising:

means for detecting an object of interest in a patient in a first x-ray image;

means for determining an area of interest based on a predicted motion of the object of interest;

means for acquiring a second x-ray image based on the area of interest determined in the first x-ray image using spatial x-ray modification to control an x-ray to pass through a portion of the patient corresponding to the area of interest, the second x-ray image having a reduced imaging size of the patient relative to the first x-ray image;

means for determining whether the object of interest was successfully detected in the second x-ray image; and

means for generating an x-ray scatter image of a region beyond the area of interest determined in the first x-ray image in response to a determination that the object of interest was not successfully detected in the second x-ray image.

30. The system of claim 29 , wherein the means for determining an area of interest based on a predicted motion of the object of interest comprises:

means for predicting a next location of the object of interest by predicting a motion of the object of interest; and

means for defining the area of interest as a region surrounding the predicted next location of the object of interest.

31. The system of claim 29 , wherein the means for detecting an object of interest in a first x-ray image comprises:

means for automatically detecting the object of interest in the first x-ray image using a probabilistic boosting tree (PBT) trained on annotated training data.

32. The system of claim 29 , further comprising:

means for detecting the object of interest in the second x-ray image;

means for determining a second area of interest based on a second predicted motion of the object of interest detected in the second x-ray image; and

means for acquiring a third x-ray image of the second area of interest using spatial x-ray modification to control an x-ray to pass through a portion of the patient corresponding to the second area of interest.

33. The system of claim 29 , further comprising:

means for detecting the object of interest in the x-ray scatter image.

34. The system of claim 33 , wherein the means for detecting the object of interest in the x-ray scatter image comprises:

means for detecting the object of interest in the x-ray scatter image using a probabilistic boosting tree (PBT) trained on annotated training data comprising x-ray scatter image data.