Non-contrast dynamic MRI myocardial perfusion analysis and visualization

• US 10,368,777 B2
• Filed: 09/14/2012
• Issued: 08/06/2019
• Est. Priority Date: 09/14/2012
• Status: Active Grant

• Current Assignee: Toshiba Medical Systems Corporation (Canon Inc.)
• Original Assignee: Toshiba Medical Systems Corporation (Canon Inc.)
• Inventors: Zhou, Xiangzhi, Miyazaki, Mitsue, Hoshino, Tsutomu
• Primary Examiner(s): Gillman, Amelie R
• Application Number:
  US13/618,001
• Publication Number:
  US 20140081125A1
• Time in Patent Office: 2,517 Days
• Field of Search: 600419
• CPC Class Codes: A61B 5/0044 for the heart
• CPC Class Codes: A61B 5/0263 using NMR
• CPC Class Codes: A61B 5/055 involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
• CPC Class Codes: A61B 5/742 using visual displays A61B5/339 takes precedence
• CPC Class Codes: G01R 33/385 using gradient magnetic field coils
• CPC Class Codes: G01R 33/4833 using spatially selective excitation of the volume of interest, e.g. selecting non-orthogonal or inclined slices
• CPC Class Codes: G01R 33/5608 Data processing and visualization specially adapted for MR, e.g. for feature analysis and pattern recognition on the basis of measured MR data, segmentation of measured MR data, edge contour detection on the basis of measured MR data, for enhancing measured MR data in terms of signal-to-noise ratio by means of noise filtering or apodization, for enhancing measured MR data in terms of resolution by means for deblurring, windowing, zero filling, or generation of gray-scaled images, colour-coded images or images displaying vectors instead of pixels image data processing or generation, in general G06T
• CPC Class Codes: G01R 33/56366 Perfusion imaging

Abstract

Black blood time to inversion (BBTI) tag-on and tag-off images acquired by magnetic resonance imaging (MRI) are analyzed to produce difference magnitude 3D images as a function of time (BBTI values) representing blood perfusion in a region of interest (ROI). Perfusion data of the ROI having values which are different for normal and abnormal myocardial tissues are displayed for plural slices of a 3D image and for plural BBTI values in a single display panel.

Claims

1. A computerized system for displaying images acquired by magnetic resonance (MR) imaging, said system comprising:

at least one computer processor coupled to associated memory, display and input/output ports and configured to;

(a) acquire separate 3D multi-slice non-contrast reconstructed MR images of blood perfusion within left ventricle (LV) myocardium for each of plural inversion time (TI) intervals during a same cardiac phase;

(b) perform segmentation of each of the 3D multi-slice non-contrast reconstructed MR images; and

(c) display, to an operator of said system, separate LV slice reconstructed images of blood perfusion at each of plural slice positions from each of the 3D multi-slice non-contrast reconstructed MR images on which the segmentation has been performed, for each of said plural TI time intervals, in sequences of respectively corresponding LV slice reconstructed images of blood perfusion at each of said plural slice positions, as a function of the TI time intervals, simultaneously in a single display panel, in a color map format for each segment resulting from the segmentation.

2. A computerized system for analyzing inversion time (TI) tag-on and tag-off non-contrast images acquired by magnetic resonance imaging (MM) during a same cardiac phase, said system comprising:

at least one computer processor coupled to associated memory, display and input/output ports and configured to;

(a) subtract tag-off and tag-on images, each of the tag-off images and each of the tag-on images having been acquired during a same cardiac phase in relation to respective electrocardiogram (ECG) signals and using multi-slice non-contrast imaging, on a pixel-by-pixel basis using both real and imaginary parts of complex-valued acquired MM signals for each of plural TI time intervals to produce difference magnitude 3D images for a 3D volume as a function of the TI time intervals, said difference magnitude 3D images representing blood perfusion in left ventricle (LV) myocardium tissue with blood perfusion data values which are different for normal and abnormal myocardial tissues; and

(b) display separate LV slice reconstructed MR images of said difference magnitude 3D images for each of plural different slices of said 3D volume at plural of the TI time intervals simultaneously in a single display panel.

3. The system of claim 2, wherein said at least one computer processor is further configured to effectimage registration between said difference magnitude 3D images by separately registering, using both said tag-on and said tag-off images, (a) tag-on and/or tag-off images having positive contrast blood where myocardial MR signal intensity is greater than that of left ventricle chamber blood and (b) tag-on and/or tag-off images having negative contrast blood where myocardial MR signal intensity is less than that of left ventricle chamber blood, and

recording corresponding registration pixel shifts in the difference magnitude 3D images required for the separate registrations.

4. The system of claim 2, wherein said at least one computer processor is further configured to

perform rigid or non-rigid 3D image registration by separately registering (a) first images, from said tag-off and tag-on images, that have positive contrast blood where myocardial MR signal intensity is greater than that of left ventricle chamber blood and (b) second images, from said tag-off and tag-on images, that have negative contrast blood where myocardial MR signal intensity is less than that of left ventricle chamber blood, and

recording corresponding registration pixel shifts in the difference magnitude 3D images required for the separate registrations.

5. The system of claim 4, wherein said at least one computer processor is further configured to accept manual user alignment inputs to effect said 3D image registration.

6. The system of claim 2, wherein said at least one computer processor is further configured to

effect segmentation of said difference magnitude 3D images depicting blood perfusion in imaged LV myocardial tissue, wherein said LV slice reconstructed MR images that are displayed as a function of TI time intervals simultaneously in the single display panel and are selected from:

(i) LV slice reconstructed MR images prior to said segmentation; and

(ii) LV slice reconstructed MR images after said segmentation.

7. The system of claim 2 wherein said at least one computer processor is configured to segment slices of said difference magnitude 3D images depicting blood perfusion data values in the LV myocardium and to display said blood perfusion data values of the difference magnitude 3D images in at least one bull'"'"'s eye configuration showing at least one of:

(a) a segmented slice of LV myocardium having blood perfusion data values plotted along concentric circles for each of the TI time intervals plotted along a radial direction; and

(b) for one of the TI time intervals, LV myocardium blood perfusion data values plotted along concentric circles for each of different slices of the LV myocardium plotted along a radial direction.

8. The system of claim 2, wherein said at least one computer processor is further configured (A) to generate histograms of said tag-on and tag-off images used to create said difference magnitude 3D images and (B) to subtract said histograms of said tag-on and tag-off images from one another with the subtraction results providing data representing susceptibility error and/or error caused by a tagging pulse affecting the tag-on and tag-off images during image data acquisition processes.

9. The system of claim 2, wherein said at least one computer processor is further configured to (A) perform coronary artery territory segmentation of said difference magnitude 3D images to produce segmented image portions of the left ventricle myocardium blood perfusion data by using a groove between left and right ventricle myocardium as a start point for segmentation and then automatically segmenting said left ventricle myocardium spatially away from said start point, (B) generate an average image blood perfusion data value for each said segmented portion of the left ventricle myocardium, and (C) display said segmented portions of left ventricle myocardium blood perfusion data with visual differences representing their respectively associated different average image blood perfusion data values.

10. The system of claim 2, wherein said at least one computer processor is further configured to

(A) generate from said difference magnitude 3D images at least one curve as a function of the TI time intervals for each of said plural slices, and

(B) display the resulting plural curves in the single display panel, said plural curves being depicted as at least one of:

(i) an array of blood perfusion data as a function of the TI time intervals before curve fitting, (ii) an array of curves fitted to blood perfusion data as a function of the TI time intervals, and (iii) quantified curve data obtained from curves fitted to blood perfusion data as a function of the TI time intervals.

11. The system of claim 2, wherein said at least one computer processor is further configured to segment portions of a region of interest (ROI) for each of the plural different slices, and to generate a curve as a function of the TI time interval for each of the plural segmented portions for each of said plural different slices and to display, in the single panel while the separate LV slice reconstructed MR images are displayed, one of said curves corresponding to an operator selected segment of one or more of said plural different slices.

12. A computerized method for analyzing inversion time (TI) tag-on and tag-off non-contrast images acquired by magnetic resonance imaging (MM) during a same cardiac phase, said method comprising:

(a) subtracting, using at least one computer processor coupled to associated memory, display and input/output ports, tag-off and tag-on images, each of the tag-off images and each of the tag-on images having been acquired during a same cardiac phase in relation to respective electrocardiogram (ECG) signals and using 3D multi-slice non-contrast imaging, on a pixel-by-pixel basis using both real and imaginary parts of complex-valued acquired MRI signals for each of plural TI time intervals to produce difference magnitude 3D images of a 3D volume as a function of the TI time intervals, said difference magnitude 3D images representing blood perfusion in left ventricle (LV) myocardium tissue with data values which are different for normal and abnormal myocardial tissues;

(b) performing segmentation, using the at least one computer processor, of each of the difference magnitude 3D images producing respective segmented difference magnitude 3D image corresponding to each said difference magnitude 3D image; and

(c) displaying separate LV slice reconstructed MR images of plural different slices in each of said segmented difference magnitude 3D images at plural of the TI time interval values simultaneously in a single display panel, in a color map format for each segment resulting from the segmentation.

13. The method of claim 12 further comprising, effecting, using said at least one computer processor, 3D image registration between said difference magnitude 3D images by separately registering, using both said tag-on and said tag-off images, (a) tag-on and/or tag-off images having positive contrast blood where myocardial MR signal intensity is greater than that of left ventricle chamber blood, and (b) tag-on and/or tag-off images having negative contrast blood where myocardial MR signal intensity is less than that of left ventricle chamber blood, and

recording, using the at least one computer processor, corresponding registration pixel shifts in the difference magnitude 3D images required for the separate registrations.

14. The method of claim 12 further comprising, performing, using the at least one computer processor, rigid or non-rigid 3D image registration by separately registering (a) first images, from said tag-off and tag-on images, that have positive contrast blood where myocardial MR signal intensity is greater than that of left ventricle chamber blood and (b) second images, from said tag-off and tag-on images, that have negative contrast blood where myocardial MR signal intensity is less than that of left ventricle chamber blood, and

recording, using the at least one computer processor, registration pixel shifts required for the separate registrations.

15. The method of claim 14 further comprising, accepting, using the at least one computer processor, manual user alignment inputs to effect said 3D image registration.

16. The method of claim 12, wherein said performing segmentation includes using a groove between left and right ventricle myocardium as a start point, and then automatically segmenting said left ventricle myocardium starting from said start point, wherein said LV slice images that are displayed as a function of BBTI time intervals in the single display panel are selected from:

(i) LV slice images of blood perfusion data prior to said segmentation; and

(ii) LV slice images of blood perfusion data after said segmentation.

17. The method of claim 12 further comprising, using the at least one computer processor, (A) generating histograms of said tag-on and tag-off images used to create said difference magnitude 3D images and (B) subtracting said histograms of said tag-on and tag-off images from one another with the subtraction results providing data representing susceptibility error and/or error caused by a tagging pulse affecting the tag-on and/or tag-off myocardium images during image data acquisition processes.

18. The method of claim 12, wherein said performing segmentation includes producing said segments by marking a groove between left and right ventricle myocardium as a start point for segmentation and then automatically segmenting said left ventricle myocardium spatially away from said start point, and wherein the method further comprises, using the at least one computer processor, (A) generating an average image blood perfusion data value for each said segment, and (B) displaying said segments with visual differences representing their respectively associated different average image blood perfusion data values.

19. The method of claim 12 further comprising, using the at least one computer processor, (A) generating from said difference magnitude 3D images at least one curve as a function of the TI time intervals for each of said plural slices, and

(B) displaying the resulting plural curves in the single display panel, said plural curves depicting at least one of:

(i) an array of blood perfusion data as a function of the TI time intervals before curve fitting, (ii) an array of curves fitted to blood perfusion data as a function of the TI time intervals, and (iii) quantified curve data obtained from curves fitted to blood perfusion data as a function of the TI time intervals.

20. The method of claim 12 wherein the performing segmentation includes producing said segments by segmenting a region of interest (ROI) for said each of plural different slices of one of said difference magnitude 3D images, and wherein the method further comprises generating a curve as a function of the TI time intervals for each of the segments of the ROI for each of plural slices of one of said difference magnitude 3D images and displaying one of said curves corresponding to an operator selected segment of one of the displayed slices.

21. The method of claim 12, wherein the method further comprises displaying said data values of the difference magnitude 3D images in at least one bull'"'"'s eye configuration showing at least one of:

(a) a segmented slice of said LV myocardium blood perfusion data values plotted along concentric circles for each of different the TI time intervals plotted along a radial direction; and

(b) for one of the TI time intervals, LV myocardium blood perfusion data values plotted along concentric circles for each of different slices of the LV myocardium plotted along a radial direction.

22. A magnetic resonance imaging (MM) system comprising:

an MRI gantry having static and gradient magnet assemblies and at least one radio frequency (RF) coil;

MRI control circuits connected to control components within said MRI gantry and configured to effect MRI data acquisition sequences of RF and gradient magnetic pulses which elicit MM signals, to acquire and process said elicited MM signals into MR image data in conjunction (a) with use of an initial spatially-selective nuclear magnetic resonance (NMR) RF tag pulse (tag-on) in a data acquisition sub-sequence and (b) without use of an initial spatially-selective NMR RF tag pulse (tag-off) in a data acquisition sub sequence;

said MM control circuits being configured to

(A) acquire multi-dimensional MR k-space data from the elicited MRI signals using said tag-on and tag-off data acquisition sub-sequences for each of plural time to inversion (TI) time intervals during a same cardiac phase without using an injected contrast agent;

(B) reconstruct said acquired k-space data into spatial domain tag-on and tag-off MR image data, each of the tag-off MR image data and each of the tag-on MR image data having been acquired during a same cardiac phase in relation to respective electrocardiogram (ECG) signals and using 3D multi-slice imaging;

(C) subtract said reconstructed tag-off and tag-on MR image data for each of plural TI time intervals using both real and imaginary parts of complex-valued acquired MRI signals from the elicited MM signals to produce difference magnitude 3D image data as a function of time having a corresponding blood flow time and/or relative peak blood flow magnitude, or lack thereof, which differentiates between normal, ischemic and infarct tissues;

(D) perform segmentation of each of the difference magnitude 3D image data; and

(E) concurrently display plural separate left ventricle (LV) slice reconstructed MR images of said difference magnitude 3D image data on which the segmentation has been performed, for each of plural slices of a 3D volume as visually different display data values (a) for normal, ischemic and infarct tissues based on said difference magnitude 3D image data and (b) at the plural TI time intervals simultaneously in a single display panel, in a color map format for each segment resulting from the segmentation.

23. The MRI system of claim 22, wherein said MRI control circuits are further configured (A) to generate histograms of said tag-on and tag-off MR image data used to create said difference magnitude 3D image data and (B) to subtract said histograms of said tag-on and tag-off images from one another with the subtraction results providing data representing susceptibility error and/or error caused by an RF tag pulse affecting the tag-on and tag-off images during image data acquisition processes.

24. The MRI system of claim 22, wherein said performing segmentation includes performing coronary artery territory segmentation of said difference magnitude 3D image data to produce said segments of the left ventricle myocardium by using a groove between left ventricle myocardium and right ventricle myocardium as a start point for segmentation and then automatically segmenting said left ventricle myocardium spatially away from said start point,

and wherein said MM control circuits are further configured to generate an average image data value for each said segmented portion of the left ventricle myocardium, and display said segments of left ventricle myocardium with visual differences representing their respectively associated different average image data values.

25. The MRI system of claim 22, wherein said MRI control circuits are further configured to generate at least one curve as a function of the TI time intervals for each of said plural slices and to display the resulting plural curves in the single display panel.

26. The MRI system of claim 22, wherein said performing segmentation includes producing said segments by segmenting a region of interest (ROI) for each of said plural slices,

and wherein said MRI control circuits are further configured to generate a curve as a function of the TI time intervals for each of the segments of the ROI for each of plural slices of said difference magnitude 3D image data and to also display one of said curves corresponding to an operator selected segment of one of the plural slices in said single display panel.