Three dimensional localization and tracking for adaptive radiation therapy
The present disclosure relates to systems, methods, and computer-readable storage media for segmenting medical image. Embodiments of the present disclosure may locate and track a moving, three-dimensional (3D) target in a patient undergoing image-guided radiation therapy. For example, an adaptive filter model for a region of interest in the patient may be received, wherein the adaptive filter model is based on the target to be tracked. An image acquisition device may obtain a two-dimensional (2D) slice of a region of interest in the patient. A processor may then apply the adaptive filter model to the 2D slice, wherein the adaptive filter model includes an offset value. The processor may also determine a location of the target in the 2D slice based on the adaptive filter model. The processor may also estimate a potential location of the target based on the offset value. The processor may then repeat one or more of the above steps to track the moving target during image-guided radiation therapy of the patient.
1. A method to locate and track a moving, three-dimensional (3D) target in a patient undergoing image-guided radiation therapy, the method comprising:
(a) receiving an adaptive filter model for a region of interest in the patient, wherein the adaptive filter model is based on the target to be tracked;
(b) obtaining, by an image acquisition device, a two-dimensional (2D) slice of a region of interest in the patient;
(c) applying, by a processor, the adaptive filter model to the 2D slice, wherein the adaptive filter model includes an offset value;
(d) determining, by the processor, a location of the target in the 2D slice based on the adaptive filter model;
(e) estimating, by the processor, a potential location of the target based on the offset value; and
repeating, by the processor, (b) through (e) to track the moving target during image-guided radiation therapy of the patient.
2. The method of claim 1 , wherein obtaining the 2D slice is based on a predetermined acquisition protocol provided by a user.
3. The method of claim 1 , wherein obtaining the 2D slice is based on the determined location of the target.
4. The method of claim 2 , wherein the acquisition protocol includes an adjustable parameter.
5. The method of claim 4 , wherein the adjustable parameter comprises at least one of a target location parameter, a slice orientation parameter, a slice thickness parameter, and a direction of target motion in the patient.
6. The method of claim 1 , further comprising receiving a predicted 3D location of the target in the patient.
7. The method of claim 1 , wherein applying the adaptive filter model comprises applying the adaptive filter model to a plurality of locations on the 2D slice.
8. The method of claim 1 , wherein obtaining a 2D slice further comprises obtaining, by the image acquisition device, a plurality of 2D slices, wherein each 2D slice is either parallel to the target motion or orthogonal to the target motion.
9. The method of claim 1 , wherein the 2D slice is either in a frequency domain or in a spatial domain.
10. The method of claim 9 , wherein the adaptive filter model and the 2D slice are both in the frequency domain.
11. The method of claim 9 , wherein the adaptive filter model and the 2D slice are both in the spatial domain.
12. The method of claim 1 , further comprising:
generating, by the processor, a response map, wherein the response map includes a confidence score for each location of the 2D slice.
13. The method of claim 12 , wherein the confidence score indicates how well the adaptive filter model matches a particular location of the 2D slice.
14. The method of claim 13 , wherein a higher confidence score indicates a higher degree of matching between the adaptive filter model and the location on the 2D slice.
15. The method of claim 13 , wherein the confidence score provides a prediction of where the target is located within the 2D slice.
16. The method of claim 12 , wherein the processor generates one or more response maps for the 2D slice, wherein each response map corresponds to a different adaptive filter model.
17. A system for locating and tracking a moving three-dimensional (3D) target to be irradiated using an image-guided radiation therapy device according to a radiation therapy treatment plan, the system comprising:
a processor; and
a memory operatively coupled to the processor, the memory storing a plurality of medical images, a plurality of adaptive filter models, an acquisition protocol, and computer-executable instructions that when executed by the processor, causes the processor to perform the method, comprising:
(a) retrieving from memory a two-dimensional (2D) slice according to the acquisition protocol;
(b) converting, by the processor, the 2D slice to a particular processing domain;
(c) applying, by the processor, the adaptive filter model to the 2D slice, wherein the adaptive filter model includes an offset value;
(d) generating, by the processor, a response map for each adaptive filter model, wherein the response map includes a confidence score for each location of the 2D slice;
(e) identifying, by the processor, a match between the adaptive filter model and one or more locations on the 2D slice based on a high confidence score;
(f) determining, by the processor, a location of the target in the 2D slice based on the high confidence score;
(g) estimating, by the processor, a potential location of the target based on the offset value; and retrieving from memory a subsequent 2D MRI slice and repeating (b) through (g) in order to track the moving target during image-guided radiation therapy to the patient.
18. The system according to claim 17 , wherein a plurality of response maps are created, each response map corresponding to different adaptive filter model.
19. The system according to claim 17 , wherein the confidence score indicates how well the adaptive filter model matches a particular location of the 2D slice.