Image analysis – Applications – Biomedical applications
Reexamination Certificate
2000-10-24
2004-08-31
Mehta, Bhavesh M. (Department: 2625)
Image analysis
Applications
Biomedical applications
C382S173000, C382S199000, C382S256000, C382S266000, C382S219000, C382S203000, C382S106000, C382S255000, C382S308000, C382S228000, C382S288000, C382S305000, C345S215000
Reexamination Certificate
active
06785409
ABSTRACT:
BACKGROUND OF THE INVENTION
The present application relates to diagnostic medical imaging. The invention finds particular application in segmenting pixel groups within a medical image for display and use in clinical diagnostics, real time image guided surgery, therapy planning, functional MRI, and the like. It finds particular application with computations which are implemented using a “fast marching method” in the “narrow band.” It is to be appreciated however, that the present invention finds further application in segmenting or defining borders in any digitized image.
Many different types of medical imaging equipment exist today. The uses of, and the analysis upon many of these images continue to improve. For example, medical sciences are in the process of searching for locations within the human brain for traits like spoken language, reading, and moral reasoning. Currently, imaging techniques are the least intrusive and most favorable techniques available to study different regions within the brain. For example, the recent growth of “fMRI” is revolutionizing the research in the behavior of the brain while engaged in an activity. This branch of MRI is highly dependent upon the classification of different regions in the brain.
Another field of brain imaging is magnetoencephalography (MEG) and electroencephalography (EEG). These techniques have enabled researchers to understand brain activity better than ever before. While all of these brain imaging techniques provide a valuable tool for studying the functions of the brain, these techniques typically rely on the inconsistent application of human hands to localize particular regions or areas within the particular medical image. Moreover, frequently there is a large time lag between image acquisition and image segmentation which can delay evaluation, diagnosis, and/or research.
The present invention contemplates an improved method and apparatus which overcomes the above-referenced problems and others.
SUMMARY OF THE INVENTION
In one embodiment of the present invention, a method of digital image presentation includes receiving image data and fitting a curve to boundaries within the image data. At least the curve and the image data are registered and processed for human readable display.
In accordance with another aspect of the present invention, a method of segmenting a medical image includes determining a regional interest on the medical image and computing a propagation speed indicative of a rate at which contour changes. The method further includes computing an altered contour within the region of interest based on a previous contour and the propagation speed. A final contour is eventually extracted from the region of interest and displayed to a user.
In accordance with another aspect of the present invention, the region of interest includes a first set of pixels distinguishable from other sets of pixels in the medical image. The extracting a final contour step includes repetitively adjusting the altered contour until the altered contour substantially circumscribes the first set of pixels.
In accordance with another aspect of the present invention, particularly for cerebral images, the first set of pixels includes one of the set of white matter, gray matter, and cerebral spinal fluid.
In accordance with another aspect of the present invention, the first set of pixels includes pixels having a defines similarity to each other.
In accordance with another aspect of the present invention, the propagation includes curvature speed relating to curvature of the contour. The speed of the curve or contour propagation is controlled by the regional constant which one can change depending on the size of the medical organ or object to be segmented. If the object is large and if the capture range is large, which implies a large distance to cover, then the waiting factor is automatically adjusted.
In accordance with another aspect of the present invention, the propagation speed includes regional speed relating to the determined region of interest. This regional speed is computed using fuzzy characteristics of the regions. These fuzzy characteristics are the membership functions which tell the contribution of each pixel in each of the identified classes. The number of classes are user defined and can thus be changed to improve the accuracy of the segmentation process.
In accordance with another aspect of the present invention, the propagation speed includes gradient speed relating to gradient information of the medical image. This information is computed from the pixel classification process.
In accordance with another aspect of the present invention, the propagation speed includes fuzzy gradient shape, so called shape-based speed. The shape-based speed is computed using gradient methods from the pixel classified image or membership images. This shape-based speed is a combination of gradient and fuzzy characteristics.
In accordance with another aspect of the present invention, the method further includes computing a signed distance transform of the previous contour using a curve layering method in a band surrounding the contour.
In accordance with another aspect of the present invention, the computing assigned distance transform step includes determining an accepted set of pixels. A trial set and a far set of pixels is then tagged and distances of the trial set from the accepted set, and of the far set from the accepted set are calculated. The curve layering or fast marching of the pixels is accomplished by testing 32 variant combinations and solving Eikonal equations.
In accordance with another aspect of the present invention, the medical image is registered with the final contour and displayed. Additionally, is an ability to register images from multiple sources for the same organ or object of interest. A segmented contour can then be computed for both images and displayed.
One advantage of the present invention resides in the increased capture range of a contour within a medical image.
Another advantage of the present invention resides in the derivation of the propagation of the curve from the parametric contour deformable model by incorporation of fuzzy characteristics. Of the two classes of deformable models, parametric class and level set class, the present invention derives the level set class from the parametric deformable class which yields all the properties of the parametric class. Implementation using the level set class offers advantages of both the parametric and level set classes.
Another advantage of the present invention resides in the ability to handle cavities, concavities, convolutedness, splitting or merging of the contours as they grow.
Another advantage of the present invention resides in the ability to prevent shock formations such as first order, second order, third order and fourth order.
Another advantage of the present invention lies in the controlled speed with which the curve evolves depending on features within the image itself.
Yet another advantage of the present invention resides in ability to duplicate image contours consistently or apply similar processing to variant images.
Still further advantages will become apparent to those of ordinary skill in the art upon reading and understanding the following detailed description.
REFERENCES:
patent: 4731863 (1988-03-01), Sezan et al.
patent: 4952805 (1990-08-01), Tanaka
patent: 5164993 (1992-11-01), Capozzi et al.
patent: 5457754 (1995-10-01), Han et al.
patent: 5605155 (1997-02-01), Chalana et al.
patent: 5682886 (1997-11-01), Delp et al.
patent: 5768413 (1998-06-01), Levin et al.
patent: 5978510 (1999-11-01), Chung et al.
patent: 6018499 (2000-01-01), Sethian et al.
patent: 6021221 (2000-02-01), Takaha
patent: 6055337 (2000-04-01), Kim
patent: 6138045 (2000-10-01), Kupinski et al.
patent: 6201988 (2001-03-01), Bourland et al.
patent: 6282307 (2001-08-01), Armato, III et al.
patent: 6424732 (2002-07-01), Shiffman et al.
patent: 6430430 (2002-08-01), Gosche
Effect of Edge Detection, Pixel Classification, and Classification-Edge Fusion Over LV Calibration: A Two
Choobin Barry
Koninklijke Philips Electronics , N.V.
Mehta Bhavesh M.
LandOfFree
Segmentation method and apparatus for medical images using... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Segmentation method and apparatus for medical images using..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Segmentation method and apparatus for medical images using... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3337752