Systems and methods for processing echocardiogram images

1. A non-transitory machine-readable storage medium having a machine-readable program stored therein, wherein the machine-readable program, when executed on a processing system, causes the processing system to perform a method of image processing for echocardiogram images, wherein the method comprises:

(a) reading a first plurality of images of a cardiac tissue;

(b) selecting a set of user defined feature points from the first plurality of images by selecting three initial points corresponding to a left ventricular (LV) apex, a mitral annulus (MA) septal, and MA lateral positions on a first image of the plurality of images;

(c) tracking the user defined feature points through the first plurality of images, thereby yielding a plurality of tracked feature points;

(d) registering the first plurality of images to a reference image using the plurality of tracked feature points, thereby yielding a second plurality of images;

(e) cropping the second plurality of images, thereby yielding a third plurality of images, wherein the cropping is user-defined;

(f) performing a mathematical transformation on the third plurality of images from a signal space of pixel intensities to a frequency space of pixel wave numbers, thereby yielding a transformed plurality of images;

(g) computing a magnitude of each image of the transformed plurality of images, thereby yielding a set of magnitude transformed images;

(h) performing a coordinate transformation on the set of magnitude transformed images, thereby yielding a set of coordinate transformed images;

(i) quantifying displacement between pairs of the set of coordinate transformed images, thereby yielding a plurality of displacements across the set of coordinate transformed images;

(j) transforming the plurality of displacements into a Cartesian coordinate system, thereby yielding transformed plurality of displacements;

(k) using the transformed plurality of displacements, computing average strain rates of a change in the length of the cardiac tissue; and

(l) evaluating an average strain of a cardiac chamber of the cardiac tissue by integrating the average strain rates.

2. The non-transitory machine-readable storage medium of claim 1 , wherein the reference image comprises a first image of the first plurality of images, a last image of the first plurality of images, another image within the first plurality of images, or a user selected image from the first plurality of images.

3. The non-transitory machine-readable storage medium of claim 1 , wherein the registering the first plurality of images to the reference image using the plurality of tracked feature points, thereby yielding the second plurality of images comprises:

(a) calculating an affine transformation of each tracked feature point to the reference image; and

(b) applying the affine transformation to register each image to the reference image.

4. The non-transitory machine-readable storage medium of claim 1 , wherein the mathematical transformation comprises a Fourier transformation.

5. The non-transitory machine-readable storage medium of claim 1 , wherein performing the coordinate transformation on the set of magnitude transformed images includes interpolating the magnitude transformed images from an image grid onto a logarithm-scale grid.

6. The non-transitory machine-readable storage medium of claim 1 , wherein quantifying displacement between pairs of the set of coordinate transformed images includes using a Fourier transformation whereby a Fourier transformation correlation plane is obtained, wherein a position of the largest value on the correlation plane corresponds to the quantifying displacement based on how far the position deviates from a center of the correlation plane.

7. A non-transitory machine-readable storage medium having a machine-readable program stored therein, wherein the machine-readable program, when executed on a processing system, causes the processing system to perform a method of image processing for echocardiogram images, wherein the method comprises:

(a) reading a first plurality of images;

(b) selecting a set of user defined feature points from the first plurality of images by selecting three initial points corresponding to a left ventricular (LV) apex, a mitral annulus (MA) septal, and MA lateral positions on a first image of the plurality of images;

(c) tracking the user defined feature points through the plurality of images, thereby yielding a plurality of tracked feature points;

(d) registering the plurality of images to a reference image using the plurality of tracked feature points, thereby yielding a second set of plurality of images, wherein the reference image is specified from within the plurality of images;

(e) cropping the second set of plurality of images, thereby yielding a third set of plurality of images, wherein the cropping is user-defined;

(f) performing a Fourier transformation on the third set of plurality of images from a signal space of pixel intensities to a frequency space of pixel wave numbers, thereby yielding a transformed set of plurality of images;

(g) computing a magnitude of each image of the Fourier transformed set of plurality of images, thereby yielding a set of magnitude transformed images;

(h) performing a coordinate transformation on the set of magnitude transformed set of plurality of images, thereby yielding a set of coordinate transformed images;

(i) quantifying displacement between pairs of coordinate transformed images, thereby yielding a plurality of displacements across the set of coordinate transformed images;

(j) transforming the plurality of displacements into a Cartesian coordinate system, thereby yielding transformed plurality of displacements;

(k) computing average strain rates of a change in the length of a cardiac tissue using the transformed plurality of displacements; and

(l) evaluating strain by integrating the average strain rates.

8. The non-transitory machine-readable storage medium of claim 7 , wherein the reference image is one of a first frame of the first plurality of images, a last frame of the first plurality of images, or a user selected image from within the first plurality of images.

9. The non-transitory machine-readable storage medium of claim 7 , wherein the registering the first plurality of images to the reference image using the plurality of tracked feature points, thereby yielding the second plurality of images comprises:

(a) calculating a transformation of each tracked feature point to the reference image; and

(b) applying the transformation to register each image to the reference image.

10. The non-transitory machine-readable storage medium of claim 7 , wherein quantifying displacement between pairs of coordinate transformed images includes using a Fourier transformation.