HYBRID REFERENCELESS AND MULTIBASELINE PRF-SHIFT MAGNETIC RESONANCE THERMOMETRY

• US 20110175615A1
• Filed: 01/19/2011
• Published: 07/21/2011
• Est. Priority Date: 01/19/2010
• Status: Active Grant

• Current Assignee: InSightec Ltd., Board of Trustees of the Leland Stanford Junior University (Stanford University)
• Original Assignee: Board of Trustees of the Leland Stanford Junior University (Stanford University)
• Inventors: Rieke, Viola, Lustig, Michael, Pauly, Kim Butts, Grissom, William A., Levy, Yoav, Medan, Yoav
• Granted Patent: US 8,810,246 B2
• Time in Patent Office: Days
• US Class Current: 324/315
• CPC Class Codes: G01R 33/4804 Spatially selective measurement of temperature or pH
• CPC Class Codes: G01R 33/5611 Parallel magnetic resonance imaging, e.g. sensitivity encoding [SENSE], simultaneous acquisition of spatial harmonics [SMASH], unaliasing by Fourier encoding of the overlaps using the temporal dimension [UNFOLD], k-t-broad-use linear acquisition speed-up technique [k-t-BLAST], k-t-SENSE structural details of arrays of sub-coils G01R33/3415
• CPC Class Codes: G01R 33/56509 due to motion, displacement or flow, e.g. gradient moment nulling G01R33/567 takes precedence

Abstract

Proton resonance frequency shift thermometry may be improved by combining multibaseline and referenceless thermometry.

Claims

1. A method of performing proton resonance frequency (PRF) shift thermometry, the method comprising the steps of:

acquiring at least one PRF baseline image of an imaging region, each baseline image corresponding to a PRF phase background during the imaging region'"'"'s cycle of motion;

acquiring a PRF treatment image of the imaging region subsequent to a temperature change of a subregion within the imaging region;

estimating a phase background of the imaging region based, at least in part, on (i) the at least one baseline image, and (ii) a smooth extension of a phase surrounding the subregion into the subregion; and

estimating a temperature distribution within the subregion based on the treatment image and the estimated phase background.

2. The method of claim 1, wherein estimating the phase background comprises:

deriving a baseline phase from the at least one baseline image and subtracting the baseline phase from the treatment image so as to generate a residual phase; and

smoothly extending the residual phase surrounding the subregion into the subregion.

3. The method of claim 2, wherein the at least one baseline image comprises multiple baseline images, and wherein deriving the baseline phase comprises determining coefficients associated with a combination of the baseline images.

4. The method of claim 3, wherein the coefficients are determined based on similarity of the baseline images to the treatment image.

5. The method of claim 3, wherein the coefficients are determined based on proximity of the stages of the baseline images to the stage of the treatment image within at least one of a respiratory or a cardiac cycle.

6. The method of claim 3, wherein the coefficients are determined based on proximity of a location of the subregion within the baseline images to a location of the subregion within the treatment image.

7. The method of claim 2, wherein the residual phase is smoothly extended into the subregion by regression after removal of pixels corresponding to the subregion from an image of the residual phase.

8. The method of claim 7, wherein the regression is a least-square regression.

9. The method of claim 2, wherein the residual phase is smoothly extended into the subregion by a regression method not requiring removal of pixels corresponding to the subregion from an image of the residual phase.

10. The method of claim 9, wherein the regression is an iterative reweighted L₁ regression.

11. The method of claim 2, wherein smoothly extending the residual phase into the subregion comprises computing finite differences between neighboring pixels in an image of the residual phase to determine a residual phase derivative, smoothly extending the residual phase derivative into the subregion, and integrating the smoothly extended residual phase derivative to obtain a smoothly extended residual phase.

12. The method of claim 1, wherein estimating the phase background and the temperature distribution comprises minimizing a penalized likelihood cost function with respect to at least one coefficient associated with the at least one baseline image, coefficients associated with the smooth extension, and temperature-induced phase shifts within the subregion.

13. The method of claim 12, wherein the penalized likelihood cost function comprises at least one of a non-sparseness penalty or a roughness penalty for the temperature-induced phase shifts.

14. The method of claim 12, wherein the penalized likelihood cost function has the form:

15. The method of claim 1, wherein estimating the phase background comprises fitting a linear combination of the at least one baseline image and smooth pixilated basis functions to the treatment image.

16. The method of claim 15, wherein the fitting comprises iterative least-square regression.

17. The method of claim 1, wherein the phase background is based on a single baseline image selected among a plurality of baseline images.

18. The method of claim 1, wherein the phase background is based on a weighted average of multiple baseline images.

19. The method of claim 1, wherein the phase background comprises a non-linear combination of multiple baseline images.

20. The method of claim 1, wherein the treatment image and the at least one baseline image comprise k-space raw data.

21. The method of claim 20, wherein the treatment image comprises a plurality of k-space segments and each of the at least one baseline image comprises a plurality of corresponding k-space segments, and wherein at least one coefficient associated with the at least one baseline image is determined based on the treatment image for each k-space segment independently.

22. The method of claim 1, wherein the baseline and treatment images are two-dimensional.

23. The method of claim 1, wherein the baseline and treatment images are three-dimensional.

24. The method of claim 1, wherein the baseline and treatment images are acquired with multiple MRI receiver coils.

25. The method of claim 24, wherein the baseline images and the treatment images are each combined across the MRI receiver coils prior to the phase background estimation.

26. The method of claim 24, wherein the phase background is estimated independently for each MRI receiver coil, and the temperature distribution is synthesized based on weighted combinations, across receiver coils, of the treatment images and estimated phase backgrounds.

27. The method of claim 24, wherein the phase background is estimated for each MRI receiver coil while parameters associated with the estimations are shared across the receiver coils.

28. The method of claim 24, wherein the temperature distribution is estimated using a penalized-likelihood algorithm applied jointly to the baseline and treatment images acquired with the multiple MRI receiver coils.

29. The method of claim 1, further comprising registering the at least one baseline image to the treatment image and transforming the at least one baseline image based on the registration to obtain at least one adjusted baseline images, and estimating the phase background based on the at least one adjusted baseline image.

30. The method of claim 29, wherein the transformation compensates for relative in-plane motion of the imaging region between the at least one baseline image and the treatment image.

31. The method of claim 29, wherein the transformation compensates for relative out-of-plane motion of the imaging region between the at least one baseline image and the treatment image.

32. The method of claim 1, wherein at least one of the at least one baseline image is acquired in a first patient, prior to acquisition of the treatment image in a second, different patient.

33. The method of claim 1, further comprising identifying the subregion via at least one of a computer algorithm or user input.

34. A system for performing proton resonance frequency (PRF) shift thermometry, the system comprising:

(a) a magnetic resonance imaging (MRI) unit;

(b) a control module in communication with the MRI unit and configured to cause the MRI unit to;

acquire at least one PRF baseline image of an imaging region, each baseline image corresponding to a PRF phase background during the imaging region'"'"'s cycle of motion;

acquire a PRF treatment image of the imaging region subsequent to a temperature change of a subregion within the imaging region;

estimate a phase background of the imaging region based, at least in part, on (i) the at least one baseline image, and (ii) a smooth extension of a phase surrounding the subregion into the subregion; and

estimate a temperature distribution within the subregion based on the treatment image and the estimated phase background; and

(c) a display for displaying the temperature distribution.