Automatic quantification of mitral valve dynamics with real-time 3D ultrasound

Patent metadata

Field	Value
Publication number	US-9129392-B2
Application number	US-201113580833-A
Country	US
Kind code	B2
Filing date	Feb 25, 2011
Priority date	Feb 25, 2010
Publication date	Sep 8, 2015
Grant date	Sep 8, 2015

Abstract

A series of rt-3DE images of the mitral valve are quantitatively analyzed so as to enable, for example, prediction of the degree of recurrent ischemic mitral regurgitation (IMR) and comprehensive assessments of leaflet tethering and “tenting” for the entire mitral valve. In accordance with the method, first, the rt-3DE images are registered with symmetric diffeomorphism to obtain information about how the mitral valve deforms over time. Second, the mitral valve is segmented with the level sets or other known segmentation method at each time point in the cardiac cycle with minimal user interaction. Third, the information about mitral valve structure is reduced into a 3D medial model, a compact representation of shape. In other words, a volumetric segmentation of the mitral valve is condensed to a form that is amenable to clinically relevant morphometry.

First claim

The invention claimed is:

1. A processor-implemented method of quantifying dynamic movement of a mitral valve using 3 dimensional images of a mitral valve, comprising:
a processor obtaining a series of N 3-dimensional images of the mitral valve over time;
the processor registering each pair of consecutive 3-dimensional images using symmetric diffeomorphism to obtain information about how the mitral valve deforms over time;
segmenting the mitral valve in the N 3-dimensional images using landmark points on the mitral valve in the N 3-dimensional images at predetermined time points in the cardiac cycle;
the processor using the information about how the mitral valve deforms over time to fit a 3 dimensional medial model to the segmented mitral valve at each of the predetermined time points in the cardiac cycle; and
outputting the medial representations of the mitral valve as a quantification of said dynamic movement of said mitral valve over time.

2. The method of claim 1 , wherein segmenting the mitral valve comprises initializing a region of interest in the N 3-dimensional images containing the mitral valve and then segmenting the mitral valve from the region of interest.

3. The method of claim 2 , wherein segmenting the mitral valve further comprises warping segmented mitral leaflets with deformation fields obtained in said registering step at a time point in the cardiac cycle to initialize the mitral valve segmentation at other time points in the cardiac cycle and then segmenting the initialized mitral valve.

4. The method of claim 1 , further comprising the processor building a 3D medial template for initialization and the processor creating a 4D shape prior of the mitral valve to guide 3D medial template fitting to either mitral valve segmentations created in said segmenting step or directly to the 3 dimensional images so as to produce a medial representation of the mitral valve at each predetermined time point in the cardiac cycle.

5. The method of claim 4 , further comprising the processor creating a moving video image depicting the mitral valve throughout the cardiac cycle from said medial representations of the mitral valve and outputting said moving video image and a set of clinically relevant parameters derived from the medial representation of the mitral valve at each said predetermined time point in the cardiac cycle in said outputting step.

6. The method of claim 5 , wherein the clinically relevant parameters comprise at least one of tenting volume of mitral leaflets of the mitral valve, tenting index of the mitral leaflets of the mitral valve at mid-systole, regurgitant orifice area of the mitral valve, annular area of the mitral valve, and annular and leaflet displacement and velocity over time.

7. The method of claim 1 , wherein the 3-dimensional images are rt-3DE images.

8. The method of claim 1 , wherein the information about how the mitral valve deforms over time is represented as a 4D prior for the mitral valve based on the 3-dimensional medial model.

9. A computer system for quantifying dynamic movement of a mitral valve using 3 dimensional images of a mitral valve, comprising:
means for providing a series of N 3-dimensional images of the mitral valve over time;
means for registering each pair of consecutive 3-dimensional images using symmetric diffeomorphism to obtain information about how the mitral valve deforms over time;
means for enabling segmentation of the mitral valve in the N 3-dimensional images using landmark points on the mitral valve in the N 3-dimensional images at predetermined time points in the cardiac cycle;
means for using the information about how the mitral valve deforms over time to fit a 3 dimensional medial model to the segmented mitral valve at each of the predetermined time points in the cardiac cycle; and
means for outputting the medial representations of the mitral valve as a quantification of said dynamic movement of said mitral valve over time.

10. The computer system of claim 9 , wherein the means for enabling segmentation comprises means for initializing a region of interest in the N 3-dimensional images containing the mitral valve and then segmenting the mitral valve from the region of interest.

11. The computer system of claim 10 , wherein the means for enabling segmentation further comprises means for warping segmented mitral leaflets with deformation fields obtained by said registering means at a time point in the cardiac cycle to initialize the mitral valve segmentation at other time points in the cardiac cycle and means for segmenting the initialized mitral valve.

12. The computer system of claim 9 , further comprising means for building a 3D medial template for initialization and means for creating a 4D shape prior of the mitral valve to guide 3D medial template fitting to either mitral valve segmentations created by said means for enabling segmentation or directly to the 3 dimensional images so as to produce a medial representation of the mitral valve at each predetermined time point in the cardiac cycle.

13. The computer system of claim 12 , further comprising means for creating a moving video image depicting the mitral valve throughout the cardiac cycle from said medial representations of the mitral valve and said outputting means outputs said moving video image and a set of clinically relevant parameters derived from the medial representation of the mitral valve at each said predetermined time point in the cardiac cycle.

14. The computer system of claim 13 , wherein the clinically relevant parameters comprise at least one of tenting volume of mitral leaflets of the mitral valve, tenting index of the mitral leaflets of the mitral valve at mid-systole, regurgitant orifice area of the mitral valve, annular area of the mitral valve, and annular and leaflet displacement and velocity over time.

15. The computer system of claim 9 , wherein the means for providing 3-dimensional images provides rt-3DE images.

16. The computer system of claim 9 , wherein the information about how the mitral valve deforms over time is represented by the registering means as a 4D prior for the mitral valve based on the 3-dimensional medial model.

17. The method of claim 1 , wherein the 3D medial model is defined by fitting a deformable medial template to a binary segmentation of the mitral valve.

18. The computer system of claim 9 , wherein the 3 dimensional medial model is defined by fitting a deformable medial template to a binary segmentation of the mitral valve.

Assignees

• Gorman Robert C
• Pouch Alison
• Gorman Iii Joseph H
• Sehgal Chandra
• Yushkevich Paul A
• Jackson Benjamin
• Univ Pennsylvania

Inventors

• Gorman Robert C
• Pouch Alison
• Gorman Iii Joseph H
• Sehgal Chandra
• Yushkevich Paul A
• Jackson Benjamin

Classifications

• G06T2207/10076

  4D tomography; Time-sequential 3D tomography · CPC title

• G06T7/0097 Primary

  Physics · mapped topic

• G06T7/0012 Primary

  Biomedical image inspection · CPC title

• G06T2207/30048

  Heart; Cardiac · CPC title

• G06T2207/20128

  Atlas-based segmentation · CPC title