Robust classification of fat and water images from 1-point-Dixon reconstructions
View Patent ↗Dixon methods in magnetic resonance imaging generate MRI images that may contain at least two tissue components such as fat and water. Dixon methods generate images containing both tissue components and predominantly one tissue component. A first segmentation of a first tissue component is generated in a T1 weighted image. The segmentation is correlated with at least a first and a second Dixon image. The image with the highest correlation is assigned the first tissue component.
1. A method for classifying tissue type from MRI image data of an object including at least a first type and a second type of tissue, by processing a first set of MRI image data representing substantially the first type of tissue, and a combined set of MRI image data representing at least the first and the second type of tissue, comprising using a processor to:
create a segmentation from the combined set of MRI image data; and
determine a first correlation value between the segmentation and the first set of MRI image data.
2. The method as claimed in claim 1 , wherein a pixel in the segmentation has an intensity value above a threshold.
3. The method as claimed in claim 1 , further comprising:
determining a threshold correlation value;
comparing the first correlation value with the threshold correlation value; and
assigning a tissue type to the first set of MRI image data based on the comparison of the previous step.
4. The method as claimed in claim 1 , further comprising:
determining a second correlation value between the segmentation and a second set of MRI image data of the object representing substantially the second tissue.
5. The method as claimed in claim 4 , further comprising:
classifying the tissue type of the first set of MRI image data based on a relative value of the first correlation value compared to the second correlation value.
6. The method as claimed in claim 1 , wherein the first set of MRI image data is generated by applying a Dixon method.
7. The method as claimed in claim 1 , wherein the first type of tissue is a fat tissue.
8. The method as claimed in claim 1 , wherein the segmentation is a grey value based segmentation.
9. The method as claimed in claim 1 , wherein the segmentation is performed by using a method selected from the group consisting of an expectation maximization segmentation method and an Otsu threshold segmentation method.
10. The method as claimed in claim 1 , wherein a classification is applied to one or more additional objects related to the object.
11. A system for classifying tissue type from MRI image data of an object including at least a first type and a second type of tissue, by processing a first set of MRI image data representing substantially the first type of tissue, and a combined set of MRI image data representing at least the first and the second type of tissue, comprising:
a Magnetic Resonance Imaging machine that generates magnetic resonance image data;
a processor for processing the magnetic resonance data in accordance with instructions for performing the steps of:
creating a segmentation from the combined set of MRI image data; and
determining a first correlation value between the segmentation and the first set of MRI image data.
12. The system as claimed in claim 11 , wherein a pixel in the segmentation has an intensity value above a threshold.
13. The system as claimed in claim 11 , further comprising instructions to perform:
determining a threshold correlation value;
comparing the first correlation value with the threshold correlation value; and
assigning a tissue type to the first set of MRI image data based on the comparison of the previous step.
14. The system as claimed in claim 11 , further comprising instructions to perform:
determining a second correlation value between the segmentation and a second set of MRI image data of the object representing substantially the second tissue.
15. The system as claimed in claim 14 , further comprising instructions to perform:
classifying the tissue type of the first set of MRI image data based on a relative value of the first correlation value compared to the second correlation value.
16. The system as claimed in claim 11 , wherein the first set of MRI image data is generated by applying a Dixon method.
17. The system as claimed in claim 11 , wherein the first type of tissue is a fat tissue.
18. The system as claimed in claim 11 , wherein the segmentation is a grey value based segmentation.
19. The system as claimed in claim 11 , wherein the segmentation is performed by using a method selected from the group consisting of an expectation maximization segmentation method and an Otsu threshold segmentation method.
20. The system as claimed in claim 11 , wherein a classification is applied to one or more additional objects related to the object.