Automatic detection and quantification of plaque in the coronary arteries of subjects from CT scans
AssigneeBen A. Arnold
InventorsBen A. Arnold, Judd E. Reed
Life of this patent
Estimating the term…
Abstract
A method automatically detects and quantifies arterial plaque (hard plaque, soft plaque or both) in the coronary arteries of the heart from CT images. The method uses plaque definitions based on subject specific in vivo blood/muscle and fat density measurements, subject specific voxel statistical parameters and 2-D and 3-D voxel connectivity criteria to automatically identify the plaques. The locations of the major arteries are determined in a 3-D coordinate system; and the specific coordinates of the detected plaques are displayed in a plaque map for follow-up exams or ease in plaque review, editing and reporting the results.
Claims22
An automatic method to measure coronary plaque in a subject using x-ray computed tomography (CT) by acquiring a volumetric set of CT images containing voxels representing x-ray attenuation of the subject, the method comprising:
- automatically identifying a three-dimensional (3-D) region of interest (ROI) volume which includes at least a portion of the image of the heart of the subject;
- determining a 3-D coordinate system within the 3-D ROI volume;
- automatically analyzing the 3-D ROI volume to identify voxels in the volume which satisfy defined criteria;
- automatically identifying a location of at least one plaque in one coronary artery within the ROI volume using the identified voxels;
- and
- automatically determining a measure of at least one plaque of at least a portion of one coronary artery in the volume.
A method to measure coronary calcium in a living subject using x-ray computed tomography (CT) by acquiring a volumetric set of CT images containing voxels representing x-ray attenuation of the subject, the method, comprising:
- identifying a three-dimensional (3-D) image of the heart;
- identifying the location of at least one coronary artery;
- automatically analyzing the 3-D image of the heart to identify voxels above one or more threshold values;
- automatically analyzing the identified voxels to determine which voxels satisfy a prior set of connectivity criteria;
- and
- automatically determining the calcium content of at least a portion of the coronary artery.
A method to measure calcium in coronary arteries of the human body using computed tomography (CT) by acquiring CT images containing voxels representative of x-ray attenuation in the body, the method comprising:
- identifying the location of a region containing at least one artery automatically;
- identifying calcium within the artery automatically;
- and
- quantifying the calcium without operator interaction.
A system to detect and measure plaque in at least one artery of a living subject using x-ray computed tomography images, the system comprising:
- means for locating a position of the artery on at least one image;
- means for identifying three-dimensional (3-D) regions of interest (ROIs) which include the position of the artery without operator interaction;
- means for locating plaque within the artery without operator interaction;
- means for automatically locating fat, bone or fat and bone not attributable to the plaque;
- and
- means for quantifying the plaque within the 3-D ROIs not attributed to fat or bone.
A calcium scoring method of a subject using x-ray computed tomography (CT) by acquiring a plurality of images containing image elements representing x-ray attenuation of the subject, the method comprising:
- automatically determining an outline of a body region with a field of the image containing tissue;
- automatically removing bone equivalent image elements located within the field;
- automatically detecting outlines of arteries contained within the field;
- automatically scoring calcium in the CT image with the bone equivalent image elements removed;
- and
- displaying the arterial calcium scores determined from the field with detected artery outlines while excluding calcium from bone equivalent image elements.
Dependencies are inferred from the claim text.