Diffusion MRI method for generating a synthetic diffusion image at a high B-value

10698,062

15/781197

December 01, 2016

Embodiments of the present invention provide a method for generating a synthetic diffusion image. The method comprises the steps of acquiring multiple sets of initial diffusion scan data by means of diffusion weighted magnetic resonance scans at multiple initial b-values, deriving an initial diffusion image from at least part of the initial diffusion scan data, acquiring target diffusion scan data by means of a diffusion weighted magnetic resonance scan at a target b-value higher than each of the initial b-values, and generating the synthetic diffusion image by performing a cost function based reconstruction to apply a fidelity term in k space to the synthetic diffusion image based on at least the initial diffusional image, and the target diffusion scan data.

KONINKLIJKE PHILIPS N.V. (Eindhoven, NL)

Koninklijke Philips N.V. (Eindhoven, NL)

1. A method for generating a synthetic diffusion image, the method comprising: acquiring a plurality of initial diffusion scan data by means of diffusion weighted magnetic resonance scans at a plurality of initial b-values; deriving an initial diffusion image from at least part of the initial diffusion scan data; acquiring target diffusion scan data by means of a diffusion weighted magnetic resonance scan at a target b-value higher than each of the initial b-values; and generating the synthetic diffusion image by performing a cost function based reconstruction using at least the initial diffusion image and the target diffusion scan data to apply a fidelity term in k space to the synthetic diffusion image, wherein the synthetic diffusion image is at least one of a synthetic diffusion weighted imaging (DWI) image at the target b-value and a synthetic diffusion parameter map, and wherein an initial diffusion parameter map is derived from the at least part of the initial diffusion scan data to provide the initial diffusion image when the synthetic diffusion image is the synthetic diffusion parameter map, and an initial computed DWI image at the target b-value is derived from the at least part of the initial diffusion scan data to provide the initial diffusion image when the synthetic diffusion image is the synthetic diffusion DWI image at the target b-value.

2. The method of claim 1 , wherein the step of deriving the initial diffusion parameter map further comprises: reconstructing at least two initial DWI images based upon the initial diffusion scan data associated with at least two different initial b-values; calculating the initial diffusion parameter map based upon the at least two initial DWI images.

3. The method of claim 1 , wherein the step of generating the synthetic diffusion parameter map further comprises: minimizing the cost function composed of a sum of fidelity terms measuring similarities between the initial diffusion and target scan data acquired at each initial b-value and target b-value and k space data of a DWI image calculated for each corresponding initial b-value and target b-value based upon the synthetic diffusion parameter map.

4. The method of claim 1 , wherein the synthetic diffusion parameter map is one of an ADC map, a diffusion coefficient and kurtosis (DKI) map, and an intravoxel incoherent motion (IVIM) map.

5. The method of claim 1 , wherein the step of deriving the initial computed DWI image at the target b-value further comprises: reconstructing at least two initial diffusion weighted images (DWI images) based upon the initial diffusion scan data associated with at least two different initial b-values; calculating an apparent diffusion coefficient (ADC) map based upon the at least two initial DWI images; and calculating the initial computed DWI image at the target b-value based on the calculated ADC map.

6. The method of claim 1 , wherein the step of generating the synthetic diffusion DWI image at the target b-value further comprises: minimizing the cost function composed of a weighted sum of at least the fidelity term measuring similarities between the diffusion scan data acquired at the target b-value and k space data of the synthetic diffusion DWI image and a constraint term measuring similarities between the synthetic diffusion DWI image and the initial computed DWI image at the target b-value.

7. The method of claim 1 , further comprising: applying a spatial regularization term to the cost function based reconstruction to improve a signal to noise (SNR) ratio of the synthetic diffusion image.

8. A magnetic resonance imaging system for generating a synthetic diffusion image comprising: a data receiver configured to receive initial diffusion scan data acquired by means of diffusion weighted magnetic resonance scans at a plurality of initial b-values and target diffusion scan data acquired by means of a diffusion weighted magnetic resonance scan at a target b-value higher than each of the initial b-values; an initial diffusion image generator configured to derive an initial diffusion image from at least part of the initial diffusion scan data; and a synthetic diffusion image generator configured to generate the synthetic diffusion image by performing a cost function based reconstruction using at least the initial diffusion image and the target diffusion scan data to apply a fidelity term in k space to the synthetic diffusion image, wherein the synthetic diffusion image is at least one of a synthetic diffusion weighted imaging (DWI) image at the target b-value and a synthetic diffusion parameter map, and wherein an initial diffusion parameter map is derived from the at least part of the initial diffusion scan data to provide the initial diffusion image when the synthetic diffusion image is the synthetic diffusion parameter map, and an initial computed DWI image at the target b-value is derived from the at least part of the initial diffusion scan data to provide the initial diffusion image when the synthetic diffusion image is the synthetic diffusion DWI image at the target b-value.

9. The magnetic resonance imaging system of claim 8 , wherein the initial diffusion image generator further comprises: a DWI image generator configured to reconstruct at least two initial DWI images based upon the initial diffusion scan data associated with at least two different initial b-values; an initial diffusion parameter map calculator configured to calculate the initial diffusion parameter map based upon the at least two initial DWI images and output the initial diffusion parameter map to the synthetic diffusion image generator.

10. The magnetic resonance imaging system of claim 8 , wherein the synthetic diffusion parameter map is selected from one of an ADC map, a diffusion coefficient and kurtosis (DKI) map, and an intravoxel incoherent motion (IVIM) map, and wherein the synthetic diffusion image generator is further configured to minimize the cost function composed of a sum of fidelity terms measuring similarities between the initial and target diffusion scan data acquired at each initial b-value and target b-value and k space data of a DWI image calculated for each corresponding initial b-value and target b-value based upon the synthetic diffusion parameter map.

11. The magnetic resonance imaging system of claim 8 , wherein the initial diffusion image generator further comprises: a DWI image generator configured to reconstruct at least two initial DWI images based upon the initial diffusion scan data associated with at least two different initial b-values; an ADC map calculator configured to calculate the ADC map based upon the at least two initial DWI images; and a DWI image calculator configured to calculate the initial computed DWI image at the target b-value based on the calculated ADC map and output the initial computed DWI image at the target b-value to the synthetic diffusion image generator.

12. The magnetic resonance imaging system of claim 8 , wherein the synthetic diffusion image generator is further configured to minimize the cost function composed of a weighted sum of at least the fidelity term measuring similarities between the diffusion scan data acquired at the target b-value and k space data of the synthetic DWI image and a constraint term measuring similarities between the synthetic DWI image and the initial computed DWI image at the target b-value.

13. A computer program product comprising machine executable instructions for execution by a processor controlling a magnetic resonance imaging system, wherein execution of the machine executable instructions cause the processor to: acquire a plurality of initial diffusion scan data by means of diffusion weighted magnetic resonance scans at a plurality of initial b-values; derive an initial diffusion image from at least part of the initial diffusion scan data; acquire target diffusion scan data by means of a diffusion weighted magnetic resonance scan at a target b-value higher than each of the initial b-values; and generate the synthetic diffusion image by performing a cost function based reconstruction using at least the initial diffusion image and the target diffusion scan data to apply a fidelity term in k space to the synthetic diffusion image, wherein the synthetic diffusion image is at least one of a synthetic diffusion weighted imaging (DWI) image at the target b-value and a synthetic diffusion parameter map, and wherein an initial diffusion parameter map is derived from the at least part of the initial diffusion scan data to provide the initial diffusion image when the synthetic diffusion image is the synthetic diffusion parameter map, and an initial computed DWI image at the target b-value is derived from the at least part of the initial diffusion scan data to provide the initial diffusion image when the synthetic diffusion image is the synthetic diffusion DWI image at the target b-value.

7902825 March 2011 Bammer
2010/0081918 April 2010 Sugiura et al.
2011/0052031 March 2011 Feiweier et al.
2011/0280456 November 2011 Sussman
2013/0012805 January 2013 Penn
2015/0146956 May 2015 Gall et al.
2016/0054420 February 2016 Wong
2016/0084929 March 2016 Dale
2010116124 October 2010

Kawahara et al “Additional Benefit of Computed Diffusion Weighted Imaging for Detection of Hepatic Metastases at 1.5T” Clinical Imaging, 40 (2016) p. 481-485. cited by applicant
Ogura et al “Optimal B-Values for Generation of Computed High-B Value DW Images” American Journal of Roentgenology, Apr. 2016, vol. 206, No. 4, p. 713-718. cited by applicant
Blackledge et al “Computed Diffusion Weighted MR Imaging May Improve Tumor Detection” Radiology vol. 261, No. 2, Nov. 2011 p. 573-581. cited by applicant
Shafiee et al, “Apparent Ultra-High—Value Diffusion-Weighted Image Reconstruction via Hidden Conditional Random Fields” IEEE Transactions on Medical Imaging, vol. 34, No. 5, May 2015 p. 1111-1124. cited by applicant
Ueno et al “Computed diffusion-weighted imaging using 3-T magnetic resonance imaging for prostate cancer diagnosis” Eur Radiol. (2013) 23; p. 3509-3516. cited by applicant
Glaister et al Quantitative Investigative Analysis of Tumour Separability in the Prostate Gland using Ultra-high b-value Computed Diffusion Imaging; 34th Annual International Conference of the IEEE EMBS, Aug. 28-Sep. 1, 2012. cited by applicant
Grant et al “Comparison of calculated and acquired high b value diffusion-weighted imaging in prostate cancer” Abdom Imaging, (2015) 40; p. 578-586. cited by applicant
Peled et al High b-Value Apparent Diffusion-Weighted Images From CURVE-Ball DTI; Journal of Magnetic Resonance Imaging, 30: p. 243-348 (2009). cited by applicant
Rosenkrantz et al “Computed Diffusion-weighted Imaging of the Prostate at 3T: Impact on Image Quality and Tumor Detection” Eur. Radiol. (2013) 23: p. 3170-3177. cited by applicant
Moti Friedman “Improved Multi B-Value Diffusion Weighted MRI of the Body by Simultaneous Model Estimation and Image Reconstruction” Sep. 22, 2013. cited by applicant
Haldar et al “Improved Diffusion Imaging Through SNR-Enhancing Joint Reconstruction” Magnetic Resonance in Med. vol. 69, No. 1, Mar. 5, 2012 p. 277-289. cited by applicant
Shafiee Mohammad Javad et al: “Apparent Ultra-High b-Value Diffusion-Weighted Image Reconstruction via Hidden Conditional Random Fields”,IEEE Transactions on Medical Imaging, IEEE Service Center, Piscataway, NJ, US, vol. 34, No. 5, May 1, 2015. cited by applicant
Maas M.C. et al: “Contrast-to-noise ratio in extrapolated and measured high b-value diffusion weighted prostate MR images”,Proceedings of the International Society for Magnetic Resonance in Medicine, 19th Annual Meeting and Exhibition, Montreal, Quebec, May 7-13, 2011, vol. 19, No. 3066, Apr. 23, 2011. cited by applicant
Haewon Nam et al: “Distortion correction of high b-valued and high angular resolution diffusion images using terative simulated images”, Neuroimage, Academic Press, Orlando, FL, US,vol. 57, No. 3, May 3, 2011(May 3, 2011),pp. 968-978. cited by applicant

edspgr.10698062