Browse Prior Art Database

AUTOMATIC CONTOURING OF A STRUCTURE (AUTOROI)

IP.com Disclosure Number: IPCOM000212750D
Publication Date: 2011-Nov-28
Document File: 4 page(s) / 255K

Publishing Venue

The IP.com Prior Art Database

Abstract

The invention provides an algorithm and workflow for determining, based on medical imagery, a region of interest corresponding to some pathological morphological structure. The invention is a semi-automatic image segmentation and structure contouring system. The main feature of the invention allows a user to provide as much interaction as is needed to guide the system to a satisfactory contouring solution.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

AUTOMATIC CONTOURING OF A STRUCTURE (AUTOROI)

FIELD OF INVENTION

The invention generally relates to medical imaging systems and more particularly to systems structure contouring.

BACKGROUND OF THE INVENTION

It is well known that various cancers are responsible for a large number of deaths every year worldwide. When tumors are visible on three dimensional images generated by such imaging devices as Computed Tomography (CT) or Magnetic Resonance (MR), a wealth of information can be extracted for determining the best treatment and then monitoring its efficacy. However in order to extract information it is required that tumor voxels for example, be singled-out or similarly that tumors be contoured. Manual contouring is usually drab and time consuming.

Usually some pathologies are materialized on some modalities as localized blobs. For example, tumors or aneurysms are visible on CT or MR images. Determination of the contour of an object on a digital image, also known as image segmentation, is a fundamental and current problem of image processing.

 

A number of conventional techniques exist in the art. One set of conventional techniques are highly automatic techniques which provide very limited user interaction. In this technique, the system tries to determine the contour of a structure. However such an approach tends to work well on a limited and well specified number of cases, for example for cases with known shape, high contrast etc. Another set of conventional techniques are highly manual in nature and a user does most of the work, for example, the user manually segments an object on many two dimensional slices. Such an approach is very robust but is often prohibitively expensive in terms of user time.

Hence there exists a need for a semi-automatic technique to determine the exact location and extent of structures of interest to efficiently diagnose underlying pathology or monitor treatment efficacy.

BRIEF DESCRIPTION OF THE INVENTION

The invention provides a novel algorithm for tumor contouring. The main feature of the proposed technique is that it allows a user to provide as much interaction as is needed to guide the system to a satisfactory contouring output solution.

DETAILED DESCRIPTION OF THE INVENTION

The proposed technique determines, based on medical imagery, a region of interest corresponding to some pathological morphological structure. Depending on the difficulty of the case, more or less of a user’s interaction can be employed. Acco...