Image analysis – Applications – Biomedical applications
Reexamination Certificate
2000-11-22
2004-06-22
Johns, Andrew W. (Department: 2621)
Image analysis
Applications
Biomedical applications
Reexamination Certificate
active
06754376
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to image processing and more specifically to segmenting three-dimensional (3D) structures from a series of cardiac images.
Computed tomography (CT) scans provide a non-invasive method for imaging the human body. One particular area of interest is imaging the heart and the vessels contained in the heart. Doctors are especially interested in examining the coronary arteries because they provide the doctor with a measure of the patient's cardiac health.
Medical images of a subject's heart typically are comprised of several slices throughout the 3D volume of the heart. In addition, the heart may be imaged at several different instants or phases in the cardiac cycle. Thus the ensemble of all images gives a picture of the heart during the complete course of one heart beat.
It is often desirable to use the set of images to extract quantitative information about the heart motion useful in medical diagnosis. Such information includes measurement of the cardiac blood volume, the ejection fraction and the amount of wall motion. It is necessary to detect and differentiate contiguous anatomical regions of the heart to perform this measurement. Identification of contiguous regions of the same material is known as segmentation.
Doctors can use existing techniques for examining the data provided by a cardiac CT scan, including standard two and three-dimensional viewing techniques. However, if they wish to examine the three-dimensional vessel tree of the heart separately from the heart muscle and surrounding tissues, few good options exist. Tools exist that allow doctors to extract parts of the vessels from the surrounding tissues, but they either do not extract the whole vessel tree or they require significant user interaction. Often doctors must tediously segment the vessel tree by hand from either axial CT images or a 3D rendering of the image data. This may take several hours and requires the physician to manually define what is part of the vessel tree. What is needed is an automated technique for segmenting and displaying just the coronary vessel tree, which requires reduced operator interaction.
BRIEF SUMMARY OF THE INVENTION
A method for automatically extracting a three-dimensional sub-structure, for example a coronary vessel tree, from a plurality of slice images comprises the steps of depositing a seed point within a selected region of interest and segmenting the plurality of slice images responsive to the deposition of the seed point. The segmentation is performed in accordance with a plurality of predetermined classification values to extract the three-dimensional sub-structure.
REFERENCES:
patent: 5150427 (1992-09-01), Frazee et al.
patent: 5166876 (1992-11-01), Cline et al.
patent: 5671265 (1997-09-01), Andress
patent: 5832134 (1998-11-01), Avinash et al.
patent: 5898793 (1999-04-01), Karron et al.
patent: 5903664 (1999-05-01), Hartley et al.
patent: 6148095 (2000-11-01), Prause et al.
patent: 6175655 (2001-01-01), George, III et al.
patent: 6360116 (2002-03-01), Jackson, Jr. et al.
patent: 6385332 (2002-05-01), Zahalka et al.
patent: 6501848 (2002-12-01), Carroll et al.
patent: 2002/0193687 (2002-12-01), Vining et al.
Knoplioch Jerome Francois
Turek Matthew William
Vapillon Alexis
General Electric Company
Johns Andrew W.
Nakhjavar Shervin
Patnode Patrick K.
Testa Jean K.
LandOfFree
Method for automatic segmentation of medical images does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for automatic segmentation of medical images, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for automatic segmentation of medical images will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3333257