Method for adipose tissue quantification with magnetic resonance imaging

1. A method for calculating a quantitative metric of adipose tissue using a magnetic resonance imaging (MRI) system, the steps of the method comprising:

a) acquiring k-space data with the MRI system by sampling echo signals that are formed at a plurality of different echo times;

b) producing a fat-concentration map using the k-space data acquired in step a);

c) estimating a maximum fat-concentration value from the fat-concentration map produced in step b);

d) producing an adipose mask using the maximum fat-concentration value estimated in step c) and the fat-concentration map produced in step b); and

e) calculating a quantitative metric of adipose tissue using the adipose mask produced in step d).

2. The method as recited in claim 1 in which step b) includes:

i) reconstructing a water image depicting water spins in the subject and a fat image depicting fat spins in the subject by fitting the k-space data acquired in step a) to a signal model that relates the plurality of different echo signals to image intensity values for the water image and the fat image; and

ii) producing the fat-concentration map using the reconstructed water image and the reconstructed fat image.

3. The method as recited in claim 2 in which the signal model in step b)i) also accounts for a fat spectrum having multiple fat resonance peaks.

4. The method as recited in claim 1 in which step b) includes producing the fat-concentration map by fitting the k-space data acquired in step a) to a signal model that relates the plurality of different echo signals to fat-concentration values.

5. The method as recited in claim 1 in which step c) includes calculating a mean fat-concentration value over a portion of the fat-concentration map produced in step b).

6. The method as recited in claim 1 in which step d) includes thresholding the fat-concentration map produced in step b) using the maximum fat-concentration value estimated in step c).

7. The method as recited in claim 6 in which the adipose mask is produced in step d) by assigning image intensity values from pixel locations in the fat-concentration map produced in step b) to respective pixel locations in the adipose mask when the image intensity values in the fat-concentration map are at least equal to a percentage of the maximum fat-concentration value estimated in step c).

8. The method as recited in claim 7 in which the percentage of the maximum fat-concentration value is fifty percent.

9. The method as recited in claim 1 in which the quantitative metric of adipose tissue calculated in step e) is at least one of a total adipose tissue (TAT) volume, a visceral adipose tissue (VAT) volume, and a VAT-to-TAT ratio (VTR).

10. The method as recited in claim 9 in which the quantitative metric of adipose tissue is TAT volume and is calculated by determining a number of non-zero voxels in the adipose mask and multiplying the number by a volume of a voxel in the fat-concentration map.

11. The method as recited in claim 9 in which the quantitative metric of adipose tissue is VAT volume and is calculated by determining a number of non-zero voxels in the adipose mask that are associated with visceral adipose tissue and multiplying the number by a volume of a voxel in the fat-concentration map.

12. The method as recited in claim 11 in which the number of non-zero voxels in the adipose mask that are associated with visceral adipose tissue are determined using at least one of an automatic segmentation, a semi-automatic segmentation, and a manual segmentation.

13. The method as recited in claim 1 in which step b) includes removing noise from the produced fat-concentration map.

14. The method as recited in claim 13 in which noise is removed from the fat-concentration map produced in step b) by producing a noise mask and applying the produced noise mask to the fat-concentration map.

15. The method as recited in claim 14 in which the noise mask is produced using a clustering algorithm.

16. The method as recited in claim 15 in which the clustering algorithm is a k-means least-squares clustering algorithm.

17. The method as recited in claim 14 in which the noise mask is produced by identifying pixel locations in the fat-concentration map that contain at least one of background noise and air cavities.

18. The method as recited in claim 1 in which step c) includes calculating a fat concentration value automatically using histogram analysis on the fat-concentration map produced in step b).

19. A non-transitory computer readable storage medium having stored thereon a computer program comprising instructions that when executed by a processor causes the processor to:

i) estimate from a stored fat-concentration map, a maximum fat-concentration value that indicates that pixel locations in the fat-concentration map do not contain one-hundred percent fat spins;

ii) produce an adipose mask using the estimated maximum fat-concentration value and the stored fat-concentration map; and

iii) calculate a quantitative metric of adipose tissue using the adipose mask.

20. The non-transitory computer readable storage medium as recited in claim 19 in which in which the processor is further caused to estimate the maximum fat-concentration value by calculating a mean fat-concentration value over a portion of the stored fat-concentration map.

21. The non-transitory computer readable storage medium as recited in claim 19 in which in which the processor is further caused to estimate the maximum fat-concentration value by performing histogram analysis on the stored fat-concentration map.

22. The non-transitory computer readable storage medium as recited in claim 19 in which the quantitative metric of adipose tissue is at least one of a total adipose tissue (TAT) volume, a visceral adipose tissue (VAT) volume, and a VAT-to-TAT ratio (VTR).

23. The non-transitory computer readable storage medium as recited in claim 19 in which the stored fat-concentration map is a fat-concentration generated using at least one of an x-ray computed tomography system and a magnetic resonance imaging system.