Computer graphics processing and selective visual display system – Computer graphics processing – Three-dimension
Reexamination Certificate
1998-11-05
2003-04-08
Zimmerman, Mark (Department: 2671)
Computer graphics processing and selective visual display system
Computer graphics processing
Three-dimension
C345S420000, C345S421000, C345S424000, C345S581000, C128S916000, C382S154000, C382S128000, C382S131000
Reexamination Certificate
active
06545678
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to the field of imaging in general and more particularly to the imaging of tissue surfaces using volumetric data.
BACKGROUND OF THE INVENTION
Heart disease is one of the leading causes of death in the United States. An important part of diagnosing heart disease may be assessing the mechanical function of the heart. In particular, quantitative measurement of a patient's heart, including the left ventricle, may provide a basis upon which a diagnosis can be made. For example, ejection fraction, which may involve a determination of a left ventricle surface and/or volume, may be used as an indicator of a heart condition.
One method to quantitatively measure heart anatomy is to acquire moving pictures of the heart. This method may provide visualization of the heart walls and quantitative measures of the heart volume and/or surface through planimetry or other means. Volumetric measurement of the left ventricle has been done using techniques such as conductance catheters and cineangiography. Conductance catheters can measure volumes using the conductance of blood, which is proportional to the blood volume, but its accuracy may be dependent on the measurement of an offset term. Moreover, the use of catheters is an invasive technique, which can induce arrhythmia in the patient.
Volume measurements using Cineangiography may be calculated from Two-Dimensional (2D) projections of the left ventricle which may induce errors in the volume measurements. The projection error may be particularly problematic when analyzing aneurysmic hearts.
Computerized Tomography (CT) and Magnetic Resonance Imaging (MRI) have been used to analyze the heart, providing volumetric data of the left ventricle of the heart. Measurements may then be made using the volumetric data. Cardiac and respiratory gating, however, may be required due to the long acquisition times associated with CT and MRI. In particular, the acquisition time for CT and MRI volume data may span several heart beats, thereby making analysis of the heart function more difficult. Furthermore, gated CT and MRI systems may generate average quantitative measurements which may adversely affect the images and the measurements. In particular, averaging may cause spatial misregistration in the image particularly in the presence of an arrhythmia. Consequently, CT and MRI systems may not allow a functional assessment of the heart between beats.
One method of calculating the volume of a left ventricle from volumetric data is by manually tracing the endocardial border of the left ventricle of the heart in a plurality of images from a set of short axis tomographs or slices of the heart. In particular, a cardiologist may manually trace a parallel set of tomographic images of the left ventricle that include the endocardial border. A thickness may be assigned to each of the respective slices and an area of the endocardial boundary for each slice is estimated using the manual tracing for that slice. The area and thickness for each slice are used to calculate a portion of the ventricular volume represented by respective slice, and all of the respective volumes may then be summed to provide an estimated left ventricle volume.
Automatic Border Detection (ABD) may be used in some situations to automatically trace the endocardial border. Some ABD methods may treat the ventricular surface as a stack of contours and apply border detection in only two dimensions. Other methods may treat the ventricular surface as a parameterized representation. In both methods, one of the goals is to optimize the contour or maximize the boundary strength, which may be defined by the local gradient in the image. Automatic boundary tracing using these two types of ABD has been partially successful, but may not perform well on volumetric ultrasound data. In particular, ultrasound data may be noisy, have relatively low resolution, and exhibit drop out due to shadowing or poorly aligned surfaces, and therefore may be difficult to accurately process using ABD.
The tracing of the endocardial boundary may allow the generation of a surface that corresponds to the left ventricle of the heart. Primarily, triangulation and surface optimization have been used for the generation of a Three-Dimensional (3D) surface from a set of points. These techniques, however, may be computationally intensive.
Other surface reconstruction methods use a spherical surface representation and 4
th
order polynomials to interpolate the surface by triangulating a set of points local to the point of interest. This method may require a well-sampled surface or an irregular sampling structure for adequate reconstruction. Moreover, circular or line ordered samples could cause unwanted artifacts in the interpolation.
As described above, conventional methods of detecting the endocardial boundary of the left ventricle may be slow and prone to errors or may not perform well using volumetric ultrasound data. Furthermore, conventional methods of surface reconstruction of the left ventricle may be computationally intensive or require a well sampled surface to perform well. Consequently, there continues to be a need for improved methods, systems, and computer program products to provide images of tissue surfaces using volumetric data.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide improved imaging of tissue surfaces.
It is another object of the present invention to allow a reduction in the time required to generate a 3D view of the surface of a tissue.
It is another object of the present invention to provide improved methods, systems, and computer program products for generating images of a surface of the left ventricle of the heart.
These and other objects are provided by accepting manual traces of boundaries of a tissue in selected tomographic images from a user. A tissue surface is reconstructed based on the input from the user. The user selects tomographic images and traces the tissue boundaries therein to further define the reconstructed surface.
In one aspect of the present invention, an initial estimate of the tissue surface is provided using a predetermined shape and modifying the shape as the boundaries are traced, thereby further refining a 3D view of the tissue surface. For example, a cylindrical shape may be used when generating a 3D view of left ventricle of the heart. The generation of the 3D view of the tissue surface is accomplished through the use of computer graphics technology.
In particular, a plurality of tomographic images including the tissue surface are selected from volumetric data to provide a plurality of selected tomographic images, each of the selected tomographic images having a different orientation with respect to the tissue surface. Manual input of data is accepted to define a boundary of the tissue surface in each of the selected tomographic images to provide respective manual traces of the boundary of the tissue surface. A 3D view of the tissue surface is generated based on the manual traces of the boundary of the tissue surface. The volumetric data may be 3D data generated by an imaging modality, such as ultrasound, CT, or MRI, or stored in a medium for analysis at a later time, such as an optical disk.
In another aspect of the present invention, a composite image is displayed. The composite image includes a 3D view of the tissue surface and a selected tomographic image including the tissue surface from the volumetric data. The 3D view is generated using the manual traces of the boundary of the tissue surface in each of a plurality of the selected tomographic images. Displaying the 3D view in conjunction with a tomographic image may orient the user with respect to the tissue surface being viewed, thereby allowing more efficient reconstruction of the tissue surface by the user.
In another aspect of the present invention, new tomographic images are selected by traversing the rendered view of the tissue surface using a sweep and a turn position. Manual input-data is accepted to define a boundary of the tis
Duke University
Myers Bigel & Sibley & Sajovec
Nguyen Kimbinh T.
Zimmerman Mark
LandOfFree
Methods, systems, and computer program products for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods, systems, and computer program products for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods, systems, and computer program products for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3022157