Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Method and System for Automatic Detection of Types of Hypertrophic Cardiomyopathy in Short-Axis Cardiac MRI

IP.com Disclosure Number: IPCOM000238395D
Publication Date: 2014-Aug-22
Document File: 5 page(s) / 181K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method and system is disclosed for automatically detecting hypertrophy and sub-types of hypertrophy in a cardiac Magnetic resonance imaging (MRI).

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 5

Method and System for Automatic Detection of Types of Hypertrophic Cardiomyopathy in Short -

-Axis Cardiac MRI

Axis Cardiac MRI

Disclosed is a method and system for detecting hypertrophy and sub -types of hypertrophy in an automated manner in a cardiac Magnetic Resonance Imaging (MRI). The method and system automatically detects hypertrophy areas by means of combining segmentation and thickness measurement methodologies . The sub-types of hypertrophy are illustrated in Fig. 1.


[1]

As shown in fig. 1, the sub-types are A: Normal heart, B: asymmetric (septal) Hypertrophic Cardiomyopathy (HCM), C: asymmetric (septal) HCM, D: apical HCM, E: symmetric HCM (concentric HCM), F: midventricular HCM, G: masslike HCM and H: noncontiguous HCM. Arrowheads in fig. 1 indicate drawings of the various phenotypes (sub-types) of HCM showing the areas of hypertrophy.

The main components of the method and system are illustrated in Fig . 2

Figure 1

1


Page 02 of 5

Figure2

As illustrated in fig. 2, the method and system automatically segments the left ventricle . In a scenario, an endocardial boundary, also known as inner wall is extracted and then moved to epicardial boundary using the endocardial boundary . For each patient, a center slice is chosen to compute a circular Region of Interest (ROI) to be applied on all slices. In each slice, a multi-threshold is applied to find a blood pool. The inner wall boundary is obtained by smoothing the convex hull of the blood pool . Thereafter, the inner wall boundary is used in computing the outer wall boundary. A cylinder is computed to include the wall. The centroid of the cylinder is the inner wall. The mean value of the interior of an inner wall radius is used as the radius of inner circle in the cylinder. The radius in the outer circle is 30mm bigger than the inner circle. The cylinder ROI is converted to a rectangular ROI by converting a Cartesian coordinate system to a polar coordinate system. The wall tissues are expected to be a black strap on the top. Further, the region growth is used to extract the outer wall boundary in the rectangular ROI. Here, the outer wall boundary extracted from the polar coordinates is

2


Page 03 of 5

converted to Cartesian coordinates in order to identify the outer wall boundary in an original image.

...