Image analysis – Applications – 3-d or stereo imaging analysis
Reexamination Certificate
1999-02-11
2002-06-04
Au, Amelia M. (Department: 2623)
Image analysis
Applications
3-d or stereo imaging analysis
C382S218000, C345S419000
Reexamination Certificate
active
06400841
ABSTRACT:
BACKGROUND OF THE INVENTION
In medical imaging, such as computed x-ray tomography, magnetic resonance imaging, ultrasound imaging or nuclear medicine, data may be acquired over a three-dimensional volume. The data may be projected to one or more two-dimensional images for viewing by a physician by using a rendering engine. The rendering engine uses a rendering algorithm, such as ray tracing, which accepts the three-dimensional data and various operation settings, such as sampling resolution, windowing, and the like, to produce the two-dimensional images.
Often a number of two-dimensional projections must be produced on a very rapid basis to provide cinematographic-type image rotation such as assists the physician's understanding of the three-dimensional structure. Accordingly, it is desirable that the selection of the rendering algorithm and its operating settings be adjusted so as to provide the very best image in the shortest possible processing time. Generally, the operation settings will significantly affect imaging speed. Different rendering engines may also provide better rendering for particular types of data.
For this reason, it is desirable to evaluate particular rendering algorithms and their operation settings. Currently, such evaluation is done by subjectively evaluating clinical images by human operators while operation settings are adjusted on essentially a trial and error basis. This practice is slow, expensive and error prone.
BRIEF SUMMARY OF THE INVENTION
The present invention quantifies the previously subjective evaluation of images and thereby allows automated evaluation of rendering systems and operation parameters. This quantitative technique first identifies a “gold standard” of parameter settings using a simplified synthetic three-dimensional data set that yields an unambiguous optimal value. Total image energy may be used to quantify the image quality. The “gold standard” set of parameters is then applied to actual clinical data to produce a “gold standard image” which may be compared to different images of the same data using different parameters. The comparison may be made using a simple mathematical measure such as root mean square (RMS) error. A function may be fit to the data points of the different parameters and the image energy using regression or other techniques such as the statistical technique “Design of Experiment” so as to provide a quick means of evaluating trade offs between imaging parameters.
Specifically, the method provides a method of evaluating three-dimensional rendering systems with actual clinical data. The method includes the steps of collecting three-dimensional image data and identifying a set of parameters (being either or both of the rendering algorithm and the operation parameters) affecting an image quality of the two-dimensional images produced by the rendering system. A set of two-dimensional images are generated using the rendering system operating on the three-dimensional image data using different parameters of the set of parameters. For each two-dimensional image, a mathematical index, such as image energy, is produced related to the image quality of the two-dimensional image. Finally, the functional relationship between the mathematical index and the parameters is established so as to allow rapid evaluation of the affect of a change in parameter on image quality.
Thus it is one object of the invention to allow automated evaluation of the functional relationship between rendering system parameters and the produced images.
The method may include the step of generating synthetic three-dimensional image data, and the mathematical index may be generated by comparing each two-dimensional image to an optimized image. The optimized image may be generated using parameters determined with the synthetic three-dimensional image data.
Thus it is another objection of the invention to provide a quantitative benchmark of image quality for actual clinical images. By developing an optimized set of parameters using a simplified test data set, those same parameters may be applied to complex clinical data to produce a benchmark image against which other images may be compared.
The synthetic three-dimensional image data may be constructed to have no overlapping data elements in projections to the two-dimensional images.
Thus it is another object of the invention to provide for a test image data set that will provide high correlation between image energy and image quality allowing for automated image evaluation.
The parameters may be operation settings of the rendering engine including interpolation type, projection angle, sampling distance, ray density, aspect ratio or window size, or may be different rendering engines altogether.
Thus, it is yet another object of the invention to provide a method of evaluating not only the settings of rendering engines, but the particular rendering engine type as well. Such a system may allow selection of rendering engines that are particularly appropriate for different types of imaging modality.
The method may include the additional steps of producing, for each two-dimensional image, a time-cost value related to the selected parameters and indicating the time required to generate the image. A functional relationship may be fit to the time-cost values so as to allow rapid evaluation of the affect of a change in parameter on time-cost.
Thus it is another object of the invention to allow image quality to not only be compared against the imaging parameters, but also against processing time needed to produce images with those parameters.
The foregoing and other objects and advantages of the invention will appear from the following description. In the description, reference is made to the accompanying drawings which form a part hereof and in which there is shown, by way of illustration, a preferred embodiment of the invention. Such embodiment does not necessarily represent the full scope of the invention, however, and reference must be made to the claims herein for interpreting the scope of the invention.
REFERENCES:
patent: 5986662 (1999-11-01), Argiro et al.
patent: 6072497 (2000-06-01), Lichtenbelt et al.
patent: 6072498 (2000-06-01), Brittain et al.
patent: 6091422 (2000-07-01), Quaknine et al.
patent: 6222937 (2001-04-01), Cohen et al.
Rusinek et al, Quantitative and Qualitative Comparison of Volumetric and Surface Rendering Techniques, Apr. 1991, IEEE ISBN: 0018-9499, vol. 38, Issue 2, pp. 659-662.
Au Amelia M.
Dastouri Mehrdad
General Electric Company
Quarles & Brady LLP
LandOfFree
Method for evaluating three-dimensional rendering systems 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 evaluating three-dimensional rendering systems, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for evaluating three-dimensional rendering systems will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2963495