Super resolution magnetic resonance (MR) images in MR guided radiotherapy
A computer implemented method of treatment targeting includes receiving magnetic resonance (MR) images of a subject including a target region, generating at least one contour of at least one surrogate element apart from the target region in the MR images, and determining a location of the target region in each of the MR images based on a location of the at least one contour in the MR images.
1. A treatment targeting system comprising:
a computing device including a processor and a memory, the memory storing instructions that when executed by the processor cause the processor to:
receive magnetic resonance (MR) images of a subject including a target region;
generate at least one contour of at least one surrogate element apart from the target region in the MR images; and
determine a location of the target region in each of the MR images based on a location of the at least one contour in the MR images.
2. The treatment targeting system of claim 1 , further comprising an MR imaging (MRI) device coupled to the computing device to generate the MR images of the subject.
3. The treatment targeting system of claim 2 , further comprising a radiotherapy delivery device coupled to the treatment targeting system, wherein the target region includes a tumor and the instructions further cause the processor to operate the radiotherapy delivery device to direct radiotherapy at the target region based on the determined location of the target region in each of the MR images.
4. The treatment targeting system of claim 1 , wherein the instructions cause the processor to generate at least one contour of at least one surrogate element by generating a first contour of a first surrogate element and generating a second contour of a second surrogate element.
5. The treatment targeting system of claim 4 , wherein the instructions cause the processor to determine the location of the target region in each of the MR images based on a location of the first contour and the second contour in the MR images.
6. The treatment targeting system of claim 4 , wherein the first surrogate element comprises the subject's diaphragm, and the second surrogate element comprises an anterior abdominal surface of the subject.
7. The treatment targeting system of claim 1 , wherein the at least one surrogate element comprises the subject's diaphragm.
8. The treatment targeting system of claim 1 , wherein the at least one surrogate element comprises an anterior abdominal surface of the subject.
9. The treatment targeting system of claim 1 , wherein the instructions cause the processor to determine the location of the target region in each of the MR images based on the location of the at least one contour in the MR images using a ridge regression model based on a relationship between a center of the target area and the at least one surrogate element.
10. The treatment targeting system of claim 1 , wherein the instructions cause the processor to determine the location of the target region in each of the MR images based on the location of the at least one contour in the MR images using principal component analysis (PCA).
11. The treatment targeting system of claim 1 , wherein the received MR images are a plurality of first MR images having a first spatial resolution, and the instructions further cause the processor to:
generate a plurality of second MR images having a second spatial resolution based on the plurality of first MR images using a neural network model, the second spatial resolution higher than the first spatial resolution, the neural network model is trained with a plurality of downsampled images downsampled from a plurality of training images as inputs and the plurality of training images as outputs; and
generate the at least one contour of the at least one surrogate element and determine the location of the target region using the second MR images.
12. The treatment targeting system of claim 11 , wherein the plurality of first MR images are a plurality of cine MR images.
13. The computer implemented method of claim 1 , wherein the received MR images are a plurality of first MR images having a first spatial resolution, and the method further comprises:
generating a plurality of second MR images having a second spatial resolution based on the plurality of first MR images using a neural network model, the second spatial resolution higher than the first spatial resolution, the neural network model is trained with a plurality of downsampled images downsampled from a plurality of training images as inputs and the plurality of training images as outputs; and
generating the at least one contour of the at least one surrogate element and determine the location of the target region using the second MR images.
14. A computer implemented method of treatment targeting comprising:
receiving magnetic resonance (MR) images of a subject including a target region;
generating at least one contour of at least one surrogate element apart from the target region in the MR images; and
determining a location of the target region in each of the MR images based on a location of the at least one contour in the MR images.
15. The computer implemented method of claim 14 , further comprising generating the MR images of the subject using an MR imaging (MRI) device.
16. The computer implemented method of claim 15 , wherein the target region includes a tumor and the method further comprises directing, using a radiotherapy device, radiotherapy at the target region based on the determined location of the target region in each of the MR images.
17. The computer implemented method of claim 14 , wherein generating at least one contour of at least one surrogate element apart from the target region in the MR images comprises generating a first contour of a first surrogate element and generating a second contour of a second surrogate element.
18. The computer implemented method of claim 17 , wherein determining the location of the target region in each of the MR images based on a location of the at least one contour in the MR images comprises determining the location of the target region in each of the MR images based on a location of the first contour and the second contour in the MR images.
19. The computer implemented method of claim 14 , wherein determining the location of the target region in each of the MR images based on the location of the at least one contour in the MR images comprises determining the location of the target region in each of the MR images based on the location of the at least one contour in the MR images using a ridge regression model based on a relationship between a center of the target area and the at least one surrogate element.
20. The computer implemented method of claim 14 , wherein determining the location of the target region in each of the MR images based on the location of the at least one contour in the MR images comprises determining the location of the target region in each of the MR images based on the location of the at least one contour in the MR images using principal component analysis (PCA).