X-ray or gamma ray systems or devices – Specific application – Computerized tomography
Reexamination Certificate
2002-08-23
2004-07-20
Bruce, David V (Department: 2882)
X-ray or gamma ray systems or devices
Specific application
Computerized tomography
C378S008000, C378S901000
Reexamination Certificate
active
06765982
ABSTRACT:
The present application hereby claims priority under 35 U.S.C. §119 on German patent publication number DE 10141344.0 filed Aug. 23, 2001, the entire contents of which are hereby incorporated herein by reference.
FIELD OF THE INVENTION
The present invention generally relates to a method of producing images with different image characteristics from measured computer tomographic data. Preferably it relates to one in which, by use of convolution of the measured data with a first convolution core, which is designed to produce the first image characteristic, and subsequent back-projection, first image data B
1
(x
i
, y
i
) of a first image with the first image characteristic is calculated.
BACKGROUND OF THE INVENTION
A computer tomograph comprises, inter alia, an X-ray tube, X-ray detectors arranged in rows and a patient couch. The X-ray tube and the X-ray detectors are arranged on a gantry, which rotates around the patient couch and an examination axis running parallel to the latter. As an alternative to this, the X-ray detectors can also be arranged on a stationary detector ring around the patient couch, with only the X-ray tube moving with the gantry.
As a rule, the patient couch can be displaced along the examination axis, relative to the gantry. The X-ray tube produces a bundle of rays which widens in a fan shape in a layer plane at right angles to the examination axis. The limitation of this bundle of rays in the direction of the layer thickness is set by the size or the diameter of the focus on the target material of the X-ray tube and one or more stops arranged in the beam path of the bundle of X-rays. In the case of examinations in the layer plane, the bundle of X-rays passes through a layer of an object, for example a layer of a body of a patient who is supported on the patient couch, and strikes the X-ray detectors located opposite the X-ray tube. The angle at which the bundle of X-rays passes through the layer of the body of the patient, and, if appropriate, the position of the patient couch relative to the gantry change continuously during the recording of the image with the computer tomograph.
During the measurement with such a computer tomograph, a plurality of sets of measured data are obtained, which correspond to different projections of the respective transradiated layer. A set of projections, which were recorded at different positions of the gantry during the rotation of the gantry around the patient, is designated a scan. The computer tomograph records many projections at various positions of the X-ray source relative to the body of the patient in order to reconstruct an image which corresponds to a two-dimensional layer image of the body of the patient.
For this purpose, the measured data are firstly convoluted with a convolution core which, taking into account the physical relationships and the measurement system, produces specific image characteristics. Then, in order to reconstruct the two-dimensional layer image, it transforms the data into the Cartesian space of the image. This technique is also referred to as filtered back-projection. The convolution cores used during the convolution are drawn up on the basis of the desired image characteristic or are known for a large number of such image characteristics. These image characteristics can be, for example, the highest possible local resolution or good low-contrast detectability. In this case, by using a suitable convolution core, the desired image characteristic can be achieved in the reconstructed layer image.
In many cases, it is necessary to obtain images with different image characteristics from the measured data acquired during one measurement. For example, in many applications, an image with high local resolution is produced from the measured or raw data by use of filtered back-projection. Then, from the same measured data, again by use of filtered back-projection, this time with a different convolution core, an image with a good, low-contrast detectability is calculated and displayed. This requires storage of the measured data for the second image calculation. The storage effort is considerable in this case, on account of the large quantity of projection data sets. Furthermore, because of the large amount of measured data, the second calculation leads to a high computing effort.
SUMMARY OF THE INVENTION
An object of an embodiment of the present invention is to specify a simplified method of producing images with different image characteristics from measured computer tomographic data, which manages with a low memory requirement and a short computing time.
In the method of an embodiment of the present application, the image with the first image characteristic is calculated in a known manner by use of filtered back-projection, via convolution of the measured data with the first convolution core which gives rise to the first image characteristic. The further image with a different image characteristic is, however, not produced by renewed calculation from the raw data, but by applying a two-dimensional filter or convolution core to the image data B
1
(x
i
, y
i
) from the first image. As a result of this subsequent filtering of the first image data B
1
(x
i
, y
i
), second image data B
2
(x
i
, y
i
) are obtained which yield the second image with the different image characteristic. In this case, the two-dimensional filter is selected on the basis of the first convolution core and the desired different image characteristic.
The filter for filtering the first image data is preferably obtained from a back-transformation of the relationship of a second convolution core for measured data, the core being designed to produce the second image characteristic, to the first convolution core into the space of the first image.
As opposed to the known method of image production from the prior art, in the present method of an embodiment of the present application only filtered back-projection is required with the corresponding computing effort. Thus, the measured or raw data can be discarded following this back-projection. Intermediate storage for the production of subsequent images with other image characteristics is no longer required in this case. This leads to a considerable reduction in the storage requirement and to a reduction in the computing time for the display of a second image with a different image characteristic. In addition, the hardware outlay for the back-projection of raw data is reduced accordingly.
Instead of applying a second convolution core to the measured data, in a preferred embodiment of the present method, this second convolution core is used, in conjunction with the first convolution core, to provide or calculate a new convolution core for two-dimensional filtering in the space, that is to say in the coordinate space, of the first image. The second image is then obtained by applying this convolution core or filter to the image data from the first image, which needs a substantially lower amount of storage space than the original raw data. Of course, this filter for the space of the image does not have to be recalculated during the performance of each measurement. Instead, filters already suitable for the different image characteristics and combinations of image characteristics of the first and second image can be provided.
The present method of an embodiment of the present application can be employed particularly advantageously in computer tomographic measurements in which a first image with high local resolution and a second image with good low-contrast detectability are needed.
The calculation of the first image with high local resolution is in this case carried out in a known way with a steepening convolution core. The first image data, attained following back-projection in this case, is displayed to the operator of the computer tomograph in the conventional way. Then—or else during the image display—this image data is subjected to filtering with a filter which is obtained from a transformation of the relationship of the first convolution core to a second, smoothing convolution core i
Bruce David V
Harness & Dickey & Pierce P.L.C.
Siemens Aktiengesellschaft
LandOfFree
Method of producing images from measured computer... 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 of producing images from measured computer..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of producing images from measured computer... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3223268