Estimation of b-value in prostate magnetic resonance diffusion weighted images

1. A system of estimating b-values, the system comprising:

an electronic processor configured to

receive a set of medical images associated with a patient, wherein the set of medical images are diffusion-weighted images,

extract a set of patches from each medical image included in the set of medical images,

determine, via an estimation model trained using machine learning, a set of estimated b-values, wherein each estimated b-value is associated with a patch included in the set of patches,

determine a b-value for each of the medical images included in the set of medical images, wherein the determined b-value for each of the medical images is based on the set of estimated b-values, and

perform an image analysis function on the set of medical images based on the b-value for each of the medical images included in the set of medical images, wherein the image analysis function includes at least one selected from a group consisting of a segmentation function, a detection function, a classification function, and an arbitrary b-value DWI image generation function.

2. The system of claim 1 , wherein the set of patches are extracted from an organ of interest of the patient.

3. The system of claim 1 , wherein the estimation model is a long short-term memory model.

4. The system of claim 1 , wherein the electronic processor is configured to determine the b-value for each of the medical images by performing a statistical analysis on the set of estimated b-values.

5. The system of claim 4 , wherein the statistical analysis includes generating a histogram of the set of estimated b-values, wherein the b-value for each of the medical images is associated with a peak of the histogram.

6. The system of claim 1 , wherein each patch included in the set of patches is centered at a voxel of each of the medical images.

7. A method of estimating b-values, the method comprising:

receiving, with an electronic processor, a set of medical images associated with a patient, wherein the set of medical images are diffusion-weighted images;

extracting, with the electronic processor, a set of patches for each medical image included in the set of medical images;

determining, with the electronic processor, via an estimation model trained using machine learning, a set of estimated b-values, wherein each estimated b-value is associated with a patch included in the set of patches;

determining, with the electronic processor, a b-value for each of the medical images included in the set of medical images, wherein the determined b-value for each of the medical images is based on the set of estimated b-values; and

performing an image analysis function on the set of medical images based on the b-value for each of the medical images included in the set of medical images, wherein performing the image analysis function includes performing at least one selected from a group consisting of a segmentation function, a detection function, a classification function, and an arbitrary b-value DWI image generation function.

8. The method of claim 7 , wherein receiving the set of medical images includes receiving a medical image associated with a prostate of the patient.

9. The method of claim 7 , wherein determining the set of estimated b-values includes determining the set of estimated b-values via a long short-term memory model.

10. The method of claim 7 , wherein determining the b-value for each of the medical images includes performing a statistical analysis on the set of estimated b-values.

11. The method of claim 10 , wherein determining the b-value for each of the medical images includes generating a histogram of the set of estimated b-values and performing an analysis of the histogram to determine a peak of the histogram.

12. The method of claim 11 , wherein determining the b-value includes determining the b-value as the peak of the histogram.

13. A non-transitory, computer-readable medium storing instructions that, when executed by an electronic processor, perform a set of functions, the set of functions comprising:

receiving a set of medical images associated with a patient, wherein the set of medical images are diffusion-weighted images;

extracting a set of patches from each medical image included in the set of medical images;

determining, via an estimation model trained using machine learning, a set of estimated b-values, wherein each estimated b-value is associated with a patch included in the set of patches;

determining a b-value for each of the medical images included in the set of medical images, wherein the determined b-value for each of the medical images is based on the set of estimated b-values; and

performing an image analysis function on the set of medical images based on the b-value for each of the medical images included in the set of medical images, wherein performing the image analysis function includes performing at least one selected from a group consisting of a segmentation function, a detection function, a classification function, and an arbitrary b-value DWI image generation function.

14. The computer-readable medium of claim 13 , wherein the set of functions further comprises:

generating a histogram of the set of estimated b-values; and

performing an analysis of the histogram to determine a peak of the histogram, wherein the b-value for each of the medical images is associated with the peak of the histogram.

15. The computer-readable medium of claim 13 , wherein receiving the set of medical images includes receiving a medical image of a prostate of the patient.

16. The computer-readable medium of claim 13 , wherein determining the set of estimated b-values includes determining the set of estimated b-values via a long short-term memory model.