Method and apparatus for detecting interventricular dyssynchrony
First Claim
1. An apparatus for detecting interventricular dyssynchrony, including:
- an image segment processor segmenting a left ventricle and a right ventricle in each of a plurality of source images of a heart to form left and right ventricle segments in a corresponding plurality of segmented images, each source image including at least a portion of a cross section of the heart along a short axis plane, each source image including cross sections of the right and left ventricles of the heart in common spatial relation along the short axis plane and in common temporal relation in reference to a cardiac cycle associated with the heart, the plurality of source images and corresponding plurality of segmented images being temporally-spaced in relation to the cardiac cycle; and
an image classifier, in operative communication with the image segment processor, determining a first cross-sectional area associated with the left ventricle segment and a second cross-sectional area associated with the right ventricle segment for each segmented image, comparing, at least in part based on use of a linear regression model, the first and second cross-sectional areas for each segmented image, and classifying the heart in an interventricular dyssynchronous class or a non-dyssynchronous class, based at least in part on the comparing, by classifying the heart in the non-dyssynchronous class if the linear regression model attains statistical significance and indicates a linear relation between the first and second cross-sectional areas over the cardiac cycle and by classifying the heart in the interventricular dyssynchronous class if the linear regression model indicates there is not a linear relation between the first and second cross-sectional areas over the cardiac cycle.
2 Assignments
0 Petitions
Accused Products
Abstract
Several embodiments of a method and apparatus for detecting interventricular dyssynchrony are provided. In one embodiment, the apparatus includes an image segment processor and an image classifier. The image segment processor may segment left and right ventricles in source images of a subject'"'"'s heart to form left and right ventricle segments in segmented images. Each source image may include cross sections of the right and left ventricles in common spatial relation along the short axis plane and in common temporal relation in reference to a cardiac cycle. The source images and corresponding segmented images may be temporally-spaced in relation to the cardiac cycle. The image classifier may determine first and second cross-sectional areas associated with the left and right ventricle segments for each segmented image, compare the first and second cross-sectional areas, and classify the subject'"'"'s heart in an interventricular dyssynchronous class or a non-dyssynchronous class.
-
Citations
31 Claims
-
1. An apparatus for detecting interventricular dyssynchrony, including:
-
an image segment processor segmenting a left ventricle and a right ventricle in each of a plurality of source images of a heart to form left and right ventricle segments in a corresponding plurality of segmented images, each source image including at least a portion of a cross section of the heart along a short axis plane, each source image including cross sections of the right and left ventricles of the heart in common spatial relation along the short axis plane and in common temporal relation in reference to a cardiac cycle associated with the heart, the plurality of source images and corresponding plurality of segmented images being temporally-spaced in relation to the cardiac cycle; and an image classifier, in operative communication with the image segment processor, determining a first cross-sectional area associated with the left ventricle segment and a second cross-sectional area associated with the right ventricle segment for each segmented image, comparing, at least in part based on use of a linear regression model, the first and second cross-sectional areas for each segmented image, and classifying the heart in an interventricular dyssynchronous class or a non-dyssynchronous class, based at least in part on the comparing, by classifying the heart in the non-dyssynchronous class if the linear regression model attains statistical significance and indicates a linear relation between the first and second cross-sectional areas over the cardiac cycle and by classifying the heart in the interventricular dyssynchronous class if the linear regression model indicates there is not a linear relation between the first and second cross-sectional areas over the cardiac cycle. - View Dependent Claims (2, 3, 4, 5, 6)
-
-
7. A method of detecting interventricular dyssynchrony, including:
-
a) receiving a plurality of source images of a heart, each source image including at least a portion of a cross section of the heart along a short axis plane, each source image including cross sections of right and left ventricles of the heart in common spatial relation along the short axis plane and in common temporal relation in reference to a cardiac cycle associated with the heart, the plurality of source images being temporally-spaced in relation to the cardiac cycle; b) segmenting the left ventricle and the right ventricle in each source image to form left and right ventricle segments in a corresponding plurality of segmented images; c) determining a first cross-sectional area associated with the left ventricle segment and a second cross-sectional area associated with the right ventricle segment for each segmented image; d) comparing, at least in part based on use of a linear regression model, the first and second cross-sectional areas for each segmented image; e) classifying the heart in an interventricular dyssynchronous class or a non-dyssynchronous class based at least in part on the comparing, the classifying including; f) classifying the heart in the non-dyssynchronous class if the linear regression model attains statistical significance and indicates a linear relation between the first and second cross-sectional areas over the cardiac cycle; and g) classifying the heart in the interventricular dyssynchronous class if the linear regression model indicates there is not a linear relation between the first and second cross-sectional areas over the cardiac cycle. - View Dependent Claims (8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23)
-
-
24. A method of detecting interventricular dyssynchrony, including:
-
a) receiving one or more source images of a heart, each source image including at least a portion of a cross section of the heart along a short axis plane, the one or more source images including cross sections of right and left ventricles of the heart in common spatial relation along the short axis plane and in common temporal relation to a cardiac cycle associated with the heart; b) segmenting the left ventricle and the right ventricle from the one or more source images to form left and right ventricle segments in one or more segmented images; c) determining a first cross-sectional area associated with the left ventricle segment and a second cross-sectional area associated with the right ventricle segment from the one or more segmented images; d) comparing, at least in part based on use of a linear regression model, the first and second cross-sectional areas; and e) classifying the heart in an interventricular dyssynchronous class or a non-dyssynchronous class based at least in part on the comparing, the classifying including; f) classifying the heart in the non-dyssynchronous class if the linear regression model attains statistical significance and indicates a linear relation between the first and second cross-sectional areas over the cardiac cycle; and g) classifying the heart in the interventricular dyssynchronous class if the linear regression model indicates there is not a linear relation between the first and second cross-sectional areas over the cardiac cycle. - View Dependent Claims (25, 26, 27)
-
-
28. An apparatus for detecting interventricular dyssynchrony, including:
-
an image segment processor segmenting a left ventricle and a right ventricle from one or more source images of a heart to form left and right ventricle segments in one or more segmented images, each source image including at least a portion of a cross section of the heart along a short axis plane, the one or more source images, at least in combination, including cross sections of right and left ventricles of the heart in common spatial relation along the short axis plane and in common temporal relation in reference to a cardiac cycle associated with the heart; and an image classifier, in operative communication with the image segment processor, determining a first cross-sectional area associated with the left ventricle segment and a second cross-sectional area associated with the right ventricle segment, comparing, at least in part based on use of a linear regression model, the first and second cross-sectional areas, and classifying the heart in an interventricular dyssynchronous class or a non-dyssynchronous class based at least in part on the comparing, by classifying the heart in the non-dyssynchronous class if the linear regression model attains statistical significance and indicates a linear relation between the first and second cross-sectional areas over the cardiac cycle and by classifying the heart in the interventricular dyssynchronous class if the linear regression model indicates there is not a linear relation between the first and second cross-sectional areas over the cardiac cycle. - View Dependent Claims (29, 30, 31)
-
Specification