Surgery – Diagnostic testing – Detecting nuclear – electromagnetic – or ultrasonic radiation
Reexamination Certificate
1999-02-17
2001-03-20
Manuel, George (Department: 3737)
Surgery
Diagnostic testing
Detecting nuclear, electromagnetic, or ultrasonic radiation
C345S424000
Reexamination Certificate
active
06205350
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to a medical diagnostic method for the two-dimensional imaging of structures present in an object, which method includes the steps of
acquiring a first data set from the object in order to define voxel image values for the voxels within a three-dimensional examination zone,
generating a second data set from the first data set by a filtering operation which emphasizes the structures. The invention also relates to a system for carrying out the method as well as to an associated image processing method.
2. Description of Related Art
In this context a voxel is to be understood to mean a volume element in the three-dimensional examination zone and a pixel is an image element of the two-dimensional image. A voxel image value is a numerical value which is associated with a voxel and characterizes a physical quantity in the relevant voxel, for example the absorption of X-rays or the nuclear magnetization.
A method of this kind is already known from the publication CAR '96, Paris, pp. 260-265 (1996). The known method has a drawback in that, due to the filtering, the two-dimensional images give the observer a rather unusual image impression.
Furthermore, from the magazine “Systems and Computers in Japan”, Vol. 25, No. 2, 1994, pp. 67-80, it is known to derive a two-dimensional image for medical diagnostic purposes from a three-dimensional data set by means of a so-called MIP (Maximum Intensity Projection). A MIP takes into account the largest of the image values of the voxels projected onto the same pixel in the two-dimensional image. This method does not provide suitable imaging when the examination zone contains not only the structures to be imaged but also further structures whose image values are in the same range as the image values of the structures to be imaged. In order to solve this problem, according to the known method the zones containing the desired structures are segmented by means of an automatically formed image mask. However, such automatic segmentation, where on the basis of the image values it is decided whether a voxel belongs to a part of the structure or not, can only be done inaccurately; manual segmentation, however, would be very intricate.
SUMMARY OF THE INVENTION
It is an object of the present invention to present a method of the kind set forth in such a manner that the structures to be imaged are not modified but undesirable structures are suppressed in the two-dimensional image.
This object is achieved by taking the following steps:
determining voxels which are projected onto the same pixel in a two-dimensional image,
selecting at least one of these voxels in dependence on its voxel image value in the second data set,
generating the two-dimensional image, the image value for the pixel then being derived from the voxel image value of the selected voxel in the first data set.
By suitably choosing the filtering method, the voxels reproducing the structures to be imaged can be assigned a different (for example, larger) voxel image value than voxels containing undesirable structures which could be undesirably superposed on the structures to be reproduced in the image. According to the invention, however, such filtering is utilized only for the identification of voxels which are somewhat likely to belong to the structure to be imaged. The imaging of this structure utilizes the original image values (from the first data set) which are associated with these voxels. Consequently, the structures to be reproduced and the undesirable structures can be separated from one another, the structures to be reproduced being imaged with the original image values (or grey values) so that in respect of these structures the two-dimensional image offers the user the normal image impression.
Various possibilities exist as regards the selection of the voxels to be imaged. One possibility corresponds to a MIP, whereas another possibility corresponds to a CVP (Closest Vessel Projection). According to the CVP, the voxels of the structure (for example, of a vascular system) which are situated nearest to the observer are imaged.
In another version the background which is not covered by the structures is suppressed and reproduced, for example in black in the image. A possibility for anatomic orientation is characterized in that a projection image B
2
(u,v)) is derived from the first data set and that this projection image and the two-dimensional image B
1
(u,v)) are superposed in an additive manner.
The first data set can be acquired in various ways, for example by means of MR (MR=Magnetic Resonance) or by X-ray CT.
The invention also includes a medical diagnostic image processing method for the two-dimensional imaging of structures in an object which is defined by a first data set with voxel image values for the voxels within a three-dimensional examination zone, which method includes the following steps: a) generating a second data set from the first data set by a filtering operation which emphasizes the structures, characterized in that it also includes the following steps: b) determining voxels (x,y,z) which are projected onto the same pixel (u,v) in a two-dimensional image, c) selecting at least one of these voxels (x,y,z) in dependence on its voxel image value (F(x,y,z)) in the second data set, d) generating the two-dimensional image. The invention also includes a medical diagnostic system for the two-dimensional imaging of structures present in an object, including means for the acquisition of a first data set from the object in order to define voxel image values for the voxels within a three-dimensional examination zone, and means for generating a second data set from the first data set by a filtering operation which emphasizes the structures, characterized in that it includes a) means for determining voxels which are projected onto the same pixel in a two-dimensional image, b) means for selecting at least one of these voxels in dependence on its voxel image value in the second data set, c) means for generating a two-dimensional image, the image value (B
1
(u,v)) for the pixel (u,v) being derived from the voxel image value (I(x, y, z)) of the selected voxel in the first data set.
REFERENCES:
patent: 5630034 (1997-05-01), Oikawa et al.
patent: 6100862 (2000-08-01), Sullivan
“Computer Assisted Radiology” in CAR '96, Paris, pp. 260-265 (1996).
“Image Processing for Computer-Aided Diagnosis of Lung Cancer by CT (LSCT) in Systems and Computers in Japan”, vol. 25, No. 2 1994, pp. 67-80.
“Multi-scal line segmentation with automatic estimation of width contrast and tangential direction in 2D and 3D medical images” by C. Lorenz et al. In Computer Vision, Virtual Reality and Robotics in Medicine CVRMed and MRCAS 199. Lecture Notes in Computer Science 1205 (Springer Verlag, Berlin, 1997) p. 233-242.
Lorenz Christian
Weese Jurgen
Manuel George
U.S. Philips Corporation
Vodopia John F.
LandOfFree
Medical diagnostic method for the two-dimensional imaging of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Medical diagnostic method for the two-dimensional imaging of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Medical diagnostic method for the two-dimensional imaging of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2534697