# US 9,436,995 B2 · Discriminating between normal and abnormal left ventricles in echocardiography

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 9,436,995 B2
- Document ID: US9436995B2
- Status: Granted
- Kind: B2
- Application number: 14/262,780
- Filed: 2014-04-27
- Published: 2015-10-29
- Granted: 2016-09-06
- Earliest claimed: 2014-04-27
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: David J. Beymer; Patrick K. McNeillie; Tanveer F. Syeda-Mahmood; Quan Wang
- Web page: https://patentgrep.com/patent/US9436995B2

## CPC

- G06T 7/0014: {using an image reference approach}
- A61B 8/0883: {for diagnosis of the heart}
- A61B 8/5223: {for extracting a diagnostic or physiological parameter from medical diagnostic data (for algorithms to analyse biomedical images G06T7/0012)}
- G06F 18/22: Matching criteria, e.g. proximity measures
- G06F 18/23: Clustering techniques
- G06F 18/24: Classification techniques
- G06T 7/0012: {Biomedical image inspection}
- G06T 7/11: Region-based segmentation
- G06T 7/13: Edge detection
- G06T 7/60: Analysis of geometric attributes
- G06T 7/74: {involving reference images or patches}
- G06V 10/755: {Deformable models or variational models, e.g. snakes or active contours}
- G06V 10/7553: {based on shape, e.g. active shape models [ASM]}
- G06F 18/2411: based on the proximity to a decision surface, e.g. support vector machines
- G06T 2207/10132: Ultrasound image
- G06T 2207/20021: Dividing image into blocks, subimages or windows
- G06T 2207/20112: Image segmentation details
- G06T 2207/30048: Heart; Cardiac
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- A61B: DIAGNOSIS; SURGERY; IDENTIFICATION (usefulness limited to only animals A61D)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Embodiments of the present invention relate to discriminating between normal and abnormal left ventricles in echocardiography. In one embodiment, a method of and computer program product for discriminating between normal and abnormal left ventricles in echocardiography are provided. A first region of a first image of a heart is located. The first region depicts a chamber of the heart. A boundary of the first region is determined. A predetermined shape is compared to the boundary. A first plurality of parameters is determined that, when applied to the predetermined shape, conforms the predetermined shape to the boundary. The first plurality of parameters is provided to a classifier. An indication of normality or abnormality is received from the classifier.

## Claims

### Claim 1

A method comprising:
locating a first region of a first image of a heart by clustering, the first region depicting a chamber of the heart, the chamber having a chamber shape;
determining a boundary of the first region by contour tracing and retaining the traced contour overlapping the located first region;
comparing a predetermined shape to the boundary, the predetermined shape having been determined without reference to the first image;
determining a first plurality of parameters that, when applied to the predetermined shape, conforms the predetermined shape to the boundary and minimizes at least one error between the predetermined shape and the boundary;
providing the first plurality of parameters to a classifier; and
receiving from the classifier an indication of normality or abnormality of the chamber shape.

### Claim 2 (depends on 1)

The method of claim 1, wherein the image is an echocardiogram.

### Claim 3 (depends on 1)

The method of claim 1, wherein the image includes a four-chamber view.

### Claim 4 (depends on 1)

The method of claim 1, wherein the chamber is a left ventricle.

### Claim 5 (depends on 1)

The method of claim 1, wherein the classifier is a linear classifier.

### Claim 6 (depends on 5)

The method of claim 5, wherein the classifier is a support vector machine.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
determining a fit between the predetermined shape and the boundary; and
providing the fit to the classifier.

### Claim 8 (depends on 1)

The method of claim 1, wherein the predetermined shape is selected from the group consisting of: an ellipsoid, a spheroid and a distorted ellipse.

### Claim 9 (depends on 1)

The method of claim 1, further comprising training the classifier, wherein training the classifier comprises:
providing a second plurality of parameters to the classifier, the second plurality of parameters corresponding to a second image of a normal heart; and
providing a third plurality of parameters to the classifier, the third plurality of parameters corresponding to a third image of an abnormal heart.

### Claim 10 (depends on 1)

The method of claim 1, further comprising:
estimating the scale of the first image.

### Claim 11 (depends on 1)

The method of claim 1, further comprising:
receiving a video; and
extracting the first image from the video.

### Claim 12 (depends on 11)

The method of claim 11, further comprising:
extracting a plurality of additional images from the video.

### Claim 13 (depends on 1)

The method of claim 1, further comprising:
locating a second region the first image of the heart, the second region comprising the first region and the second region depicting four chambers of the heart.

### Claim 14 (depends on 13)

The method of claim 13, wherein locating the first region comprises:
thresholding the second region; and
determining the location of the first region within the second region.

### Claim 15 (depends on 14)

The method of claim 14, wherein determining the location of the first region within the second region comprises:
performing k-means clustering.

### Claim 16 (depends on 1)

The method of claim 1, wherein determining the first plurality of parameters comprises:
estimating a centroid of the boundary;
determining a dominant orientation of the boundary; and
using the dominant orientation to find a rotation.

### Claim 17 (depends on 1)

The method of claim 1, wherein locating the first region of the first image of the heart comprises:
determining a dominant orientation within the first image; and
performing thresholding using the first image.

### Claim 18

A computer program product for discriminating between normal and abnormal left ventricles in echocardiography, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
locate a first region of a first image of a heart by clustering, the first region depicting a chamber of the heart, the chamber having a chamber shape;
determine a boundary of the first region by contour tracing and retaining the traced contour overlapping the located first region;
compare a predetermined shape to the boundary, the predetermined shape having been determined without reference to the first image;
determine a first plurality of parameters that, when applied to the predetermined shape, conforms the predetermined shape to the boundary and minimizes at least one error between the predetermined shape and the boundary;
provide the first plurality of parameters to a classifier; and
receive from the classifier an indication of normality or abnormality of the chamber shape.

### Claim 19 (depends on 18)

The computer program product of claim 18, wherein locating the first region of the first image of the heart comprises:
determining a dominant orientation within the first image; and
performing thresholding using the first image.

### Claim 20

A device comprising:
an imager configured to collect a first image of a heart;
an image analyzer configured to receive the first image of the heart, locate a first region of the heart by clustering, and determine a boundary of the first region of the first image by contour tracing and retaining the traced contour overlapping the located first region; the first region depicting a chamber of the heart, the chamber having a chamber shape;
a geometric fitter configured to determine a first plurality of parameters that, when applied to a predetermined shape, conforms the predetermined shape to the boundary of the first region and minimizes at least one error between the predetermined shape and the boundary, the predetermined shape having been determined without reference to the first image; and
a classifier configured to receive the first plurality of parameters and provide an indication of normality or abnormality of the chamber shape.