Patent
Apparatuses, methods and systems for estimating water diffusivity and microcirculation of blood using DW-MRI data
| العنوان: | Apparatuses, methods and systems for estimating water diffusivity and microcirculation of blood using DW-MRI data |
|---|---|
| Patent Number: | 10302,723 |
| تاريخ النشر: | May 28, 2019 |
| Appl. No: | 14/940745 |
| Application Filed: | November 13, 2015 |
| مستخلص: | (“AEW”) are disclosed herein. The apparatuses, methods and systems provide a computational framework for choosing and applying the most appropriate model in different regions of a heterogeneous area on a voxel-by-voxel basis. The apparatuses, methods and systems also configure an intravoxel-incoherent-motion (IVIM) model for estimating water diffusivity and microcirculation of blood in the capillary network from DW-MRI low b-value data. In one implementation, the method uses a small number of b-values (at least 3 in the b-value range of 0-300 s/mm2, increasing the upper bound of the low b-value range by one b-value in the absence of DW-MRI signal at 300 s/mm2 and is able to synthetically generate DW-MRI data corresponding at higher b-values using the derived IVIM equation. The method also accounts for estimating non-Gaussian diffusion parameter Kapp. |
| Inventors: | Foundation for Research and Technology - Hellas (FORTH) (Heraklion, Crete, GR) |
| Assignees: | FOUNDATION FOR RESEARCH AND TECHNOLOGY —HELLAS (FORTH) (Heraklion, GR) |
| Claim: | 1. A computer processor implemented method for outputting an optimally modelled coefficient for a voxel in diffusion weighted magnetic resonance imaging, the method comprising: (a) applying, using a processor, a mono-exponential model to signal intensities for a set of b-values at a particular voxel in a region of interest (ROI) received from an imaging device; (b) estimating, using a processor, a goodness of fit of the model applied in (a) by comparing the model to the signal intensities for b-values below a predetermined b-value at the particular voxel, wherein the estimating includes: (b1) determining, using a processor, an R-square coefficient (R 2) between the mono-exponential model and the signal intensities for the set of b-values at the particular voxel according to the formula R 2 =1− SS res /SS tot where SS res is a residual sum of squares and SS tot is a total sum of squares; (b2) determining, using a processor, an adjusted R 2 coefficient according to the formula adjusted R 2 =1−(1− R 2)*(n− 1)/ n−p− 1 where n is the number of b-values used and p is the number of parameters used from the mono-exponential model; and (b3) outputting the adjusted-R 2 coefficient as a measure of the goodness of fit; (c) if the goodness of fit is less than a predetermined goodness threshold at the particular voxel, applying, using a processor, an intravoxel-incoherent-motion (IVIM) model to the signal intensities for the set of b-values at the particular voxel and determining a perfusion fraction parameter (f), true-diffusion coefficient (D) and micro-perfusion coefficient (D*) using the IVIM model; (d) outputting an apparent diffusion coefficient (ADC) determined from the mono-exponential model for the particular voxel, if (i) the goodness of fit is not less than the predetermined goodness threshold, (ii) f is equal to a lower bound of a predetermined fraction range used in applying the IVIM model in (c), or (iii) D*/D is less than 10; and (e) outputting the true diffusion (D) determined from the IVIM model for the particular voxel, if (i) the goodness of fit is less than the predetermined goodness threshold, (ii) f is not equal to the lower bound of the predetermined fraction range used in applying the IVIM model in (c), and (iii) D*/D is not less than 10. |
| Claim: | 2. The computer processor implemented method of claim 1 , wherein step (b) comprises determining, using a processor, a root-mean-square error (RMSE) between the mono-exponential model and the signal intensities for the set of b-values at the particular voxel and outputting the RMSE as a measure of the goodness of fit. |
| Claim: | 3. The computer processor implemented method of claim 1 , further comprising repeating steps (a) through (e) for a plurality of voxels in the ROI. |
| Claim: | 4. The computer processor implemented method of claim 3 , further comprising generating, using a processor, a map for each voxel in the ROI indicating whether the ADC or the true diffusion was output at steps (d) and (e), respectively. |
| Claim: | 5. The computer processor implemented method of claim 1 , wherein the predetermined b-value is 300 s/mm 2 . |
| Claim: | 6. The computer processor implemented method of claim 1 , wherein the set of b-values includes five or fewer b-values. |
| Claim: | 7. A computer processor implemented method for modeling a portion of a diffusion weighted magnetic resonance image, the method comprising: (a) determining, using a processor, a perfusion fraction parameter (f), micro-perfusion coefficient (D*) and adjusted slope (a) using a non-linear least-squares fitting technique to fit the formula S low _ b S 0 *(− a*b +(1− f)+ f *exp(− b*D *)) to signal intensities at a particular voxel in a region of interest (ROI) received from an imaging device for at least three b-values below a predetermined b-value threshold; (b) using f, D* and a determined in step (a) to determine, using a processor, a true-diffusion coefficient (D); and (c) determining, using a processor, a signal attenuation S b for a b-value greater than the predetermined b-value threshold using D determined in step (b) and the formula S b =S 0 _ diffusion*exp(− b*D) where S 0 _ diffusion is the signal intensity of true-diffusion at b=0. |
| Claim: | 8. The computer processor implemented method of claim 7 , further comprising: (d) using D determined in step (b) to determine, using a processor, a kurtosis coefficient K app from the formula: S b =S 0 *exp(− b*D+ 1/6* b 2 *D 2 *K app) |
| Claim: | 9. The computer processor implemented method of claim 8 , further comprising repeating steps (a) through (d) for a plurality of voxels in the ROI. |
| Claim: | 10. The computer processor implemented method of claim 9 , further comprising: (e) comparing, using a processor, K app determined in step (d) for each voxel to low kurtosis threshold and a high kurtosis threshold; and (f) generating, using a processor, a classification map for each voxel in the ROI indicating whether K app =0, whether K app is less than the low kurtosis threshold or whether K app is greater than the high kurtosis threshold. |
| Claim: | 11. The computer processor implemented method of claim 7 , wherein the predetermined b-value is 300 s/mm 2 . |
| Claim: | 12. The computer processor implemented method of claim 7 , wherein the at least three b-values includes five or fewer b-values. |
| Patent References Cited: | |
| Other References: | Bihan Le D. et al., “Separation of Diffusion and Perfusion in Intravoxel Incoherent Motion MR Imaging,” Radiology, Radiological Society of North America, Inc., US, vol. 168, No. 2, Jan. 1, 1988, pp. 497-505. cited by applicant Dow-Mu Koh et al.: “Intravoxel Incoherent Motion in Body Diffusion-Weighted MRI: Reality and Challenges”, American Journal of Roentgenology, vol. 196, No. 6, Jun. 1, 2011, pp. 1351-1361. cited by applicant Padhani A. R. et al., “Diffusion-Weighted Magnetic Resonance Imaging as a Cancer Biomarker: Consensus and Recommendations,” Neoplasia, Neoplasia Press, Ann Arbor, MI, US, vol. 11, No. 2, Feb. 1, 2009, pp. 102-125. cited by applicant E. E. Sigmund et al., “Intravoxel Incoherent Motion Imaging of Tumor Microenvironment in Locally Advanced Breast Cancer”, Magnetic Resonance in Medicine, vol. 65, No. 5, Feb. 1, 2011, pp. 1437-1447. cited by applicant Hildebrand Dijkstra et al., “Effects of microperfusion in hepatic diffusion weighted imaging”, European Radiology, Springer, Berlin, DE, vol. 22, No. 4, Nov. 12, 2011, pp. 891-899. cited by applicant Belmonte G. et al., “Sensitivity and specificity of prostate tumor discrimination by IVIM approximation”, Proceeding of the International Society for Magnetic Resonance in Medicine, 22nd Annual Meeting and Exhibition, Milan, Italy, May 10-16, 2014, vol. 22, Apr. 25, 2014, p. 4541. cited by applicant Freiman, Moti et al., “In Vivo assessment of optimal b-value range for perfusion-insensitive apparent diffusion coefficient imaging”, Medical Physics, AIP. Melville, NY , US, vol. 39, No. 8, Jul. 20, 2012, pp. 4832-4839. cited by applicant Freiman, Moti et al., “Reliable estimation of incoherent motion parametric maps from diffusion-weighted MRI using fusion bootstrap moves,” Medical Image Analysis, vol. 17, 2013, pp. 325-336. cited by applicant Taimouri, V et al., “Spatially constrained incoherent motion method improves diffusion-weighted MRI signal decay analysis in the liver and spleen,” Medical Physics, vol. 42, No. 4, Apr. 2015, pp. 1895-1903. cited by applicant Ogura, A. et al, “Evaluation of intravoxel incoherent motion using the Fourier analysis for prostate cancer,” European Society of Radiology, 2015, pp. 1-14. cited by applicant Voert, E. et al., “Intravoxel Incoherent Motion Protocol Evaluation and Data Quality in Normal and Malignant Liver Tissue and Comparison to the Literature,” Investigative Radiology, vol. 51, No. 2, Feb. 2016, pp. 90-99. cited by applicant International Search Report and Written Opinion, dated May 23, 2016 for PCT International Application No. PCT/EP2015/076678, filed Nov. 16, 2015. cited by applicant Lemke, A. et al., “Toward an optimal distribution of b values for intravoxel incoherent motion imaging,”Magnetic Resonance Imaging, vol. 29, No. 6, pp. 766-776, 2011, doi: 10.1016/j.mri.2011.03.004. cited by applicant Penner, A.H. et al., “Intravoxel incoherent motion model-based liver lesion characterisation from three b-value diffusion-weighted MRI,”European Radiology, vol. 23, No. 10, pp. 2773-2783, 2013, doi: 10.1007/s00330-013-2869-z. cited by applicant Liu, C. et al., “Intravoxel incoherent motion (IVIM) in evaluation of breast lesions: comparison with conventional DWI,”European Journal of Radiology, vol. 82, no. 12, pp. e782-9, 2013, doi:10.1016/j.ejrad.2013.08.006. cited by applicant Alberich-Bayarri, A. et al., “Optimisation of b-values in MR diffusion-weighted acquisitions through information theory: a mathematical justification for consensus,”ECR 2014-24th European Congress of Radiology, Mar. 6-10, 2014, Vienna, Austria. doi: 10.1594/ecr2014/B-0580. cited by applicant Cho, G.Y. et al., “Comparison of fitting methods and b-value sampling strategies for intravoxel incoherent motion in breast cancer,”Magnetic Resonance Medicine, vol. 74, No. 4, pp. 1077-1085, 2015, doi: 10.1002/mrm.25484. cited by applicant Zhang, J.L. et al., “Optimization of b-value sampling for diffusion-weighted imaging of the kidney,”Magnetic Resonance Medicine, vol. 67, No. 1, pp. 89-97, 2012, doi: 10.1002/mrm.22982. cited by applicant Dyvorne, H. et al., “Intravoxel incoherent motion diffusion imaging of the liver: optimal b-value subsampling and impact on parameter precision and reproducibility,”European Journal of Radiology, vol. 83, No. 12, pp. 2109-2113, 2014, doi: 10.1016/j.ejrad.2014.09.003. cited by applicant Jambor, I. et al., “Optimization of b-value distribution for biexponential diffusion-weighted MR imaging of normal prostate,”Journal of Magnetic Resonance Imaging, vol. 4, No. 5, pp. 1213-1222, 2014, doi: 10.1002/jmri.24271. cited by applicant Döpfert, J. et al., “Investigation of prostate cancer using diffusion-weighted intravoxel incoherent motion imaging,”Magnetic Resonance Imaging, vol. 29, No. 8, pp. 1053-1058, 2011, doi: 10.1016/j.mri.2011.06.001. cited by applicant Wurnig, M.C. et al., “Systematic analysis of the intravoxel incoherent motion threshold separating perfusion and diffusion effects: Proposal of a standardized algorithm,”Magnetic Resonance Medicine, vol. 74, No. 5, pp. 1414-1422, 2014,. doi: 10.1002/mrm.25506. cited by applicant |
| Primary Examiner: | Mudrick, Timothy A |
| Attorney, Agent or Firm: | Norton Rose Fulbright US LLP |
| رقم الانضمام: | edspgr.10302723 |
| قاعدة البيانات: | USPTO Patent Grants |
| الوصف غير متاح. |