Electricity: measuring and testing – Particle precession resonance – Using a nuclear resonance spectrometer system
Reexamination Certificate
1998-08-20
2001-07-31
Oda, Christine (Department: 2862)
Electricity: measuring and testing
Particle precession resonance
Using a nuclear resonance spectrometer system
Reexamination Certificate
active
06268729
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to a method of imaging an object by means of magnetic resonance (MR), which object is arranged in a substantially uniform, steady magnetic field, said method including the following steps:
measuring MR signals from the object by means of MR imaging pulse sequences,
filtering the measured MR signals so as to suppress shot noise, a corrected value which is related to the measured MR signals being assigned to a measuring point of a measured MR signal to be corrected, and
reconstructing an image from the filtered MR signals.
The invention also relates to an MR device for carrying out such a method.
2. Description of Related Art
A method of this kind is known from U.S. Pat. No. 5,525,906. In the context of the present patent application a k space is to be understood to mean a spatial frequency domain in which a path is followed during the measurement of the MR signals by application of gradients to the static magnetic field. The path in the k space is determined by the time integral of the gradients applied during the interval from the excitation of the nuclear spins until the instant in time at which the MR signal is measured. The measured values of the MR signals which correspond to the most important part of the path or paths produce the inverse Fourier transformed values of an image of the object. Furthermore, the term gradient is to be understood to mean temporary magnetic fields which are superposed on a steady magnetic field and cause a gradient in the steady magnetic field. MR imaging methods often utilize gradients in three respective orthogonal directions. Generally speaking, a gradient in the first direction is referred to as a read-out gradient, a gradient in the second direction as a phase encoding gradient, and a gradient in the third direction as a selection gradient. A measuring point is to be understood to mean a sampled value of the MR signal received at a sampling instant which corresponds to a position of the path in the k space.
The known method is used to suppress shot noise in the measured MR signals. In the context of the present patent application the term shot noise is to be understood to mean brief pulses which have a duration of, for example 1 microsecond or less, and a high amplitude relative to the measured MR signal. Such shot noise is caused, for example by electric discharges within the MR device or in the vicinity thereof, for example in the clothing of the staff attending the MR device. In the reconstructed MR image the shot noise present in the MR signal causes artefacts in the form of, for example a wavy pattern or a sum of wavy patterns which forms a fabric-like structure. According to the known method, a shot-noise affected value of the measuring point, which is to be corrected in relation to the measured MR signals, is replaced by the corrected value. To this end, according to the known method the energy content of a frequency band of the spectrum of the measured MR signals which is beyond the frequency band of the imaging information in the MR signal is determined. Subsequently, if the energy content exceeds a predetermined reference, the corrected value is assigned to the measuring point to be corrected. It is a drawback of this method that it is necessary to measure MR signals having a bandwidth which is larger than the bandwidth of the MR information which is to be expected on the basis of the MR imaging pulse sequence used.
SUMMARY OF THE INVENTION
It is an object of the invention to reduce the band width of the measured MR signals. To this end, the method according to the invention is characterized in that the filtering of the measured MR signals by means of the method also includes the following steps for determining the corrected value of the measuring point to be corrected:
determining a combination of a value of a parameter of the measuring point of the measured MR signal to be corrected and values of the parameter of measuring points in a vicinity of the measuring point to be corrected, and
assigning the corrected value to the measuring point to be corrected if the combination exceeds a predetermined reference. The advantage of such a method resides in the fact that the bandwidth of the signals to be measured need not be greater than the frequency content of the MR signals which is to be expected on the basis of the MR imaging pulse sequence used, so that less expensive receivers can be used for the MR signals to be measured. The invention is based on the idea that a statistical distribution of MR signals corresponding to a large part of the k space can be first-order approximated by a Gaussian distribution of white noise. As a result, a measuring point to be corrected has very likely been influenced by the shot noise if said combination exceeds the predetermined reference. In that case, for example a value zero is assigned to the measuring point to be corrected. Even though this step introduces an error in the value of the measured MR signal, it has been found that the effect of this error in the image is less disturbing than the effect exerted on the image by shot noise in the MR signal.
A special version of the method according to the invention is characterized in that said combination is determined by the ratio of the value of the parameter of the measuring point to be corrected to a statistical magnitude of the values of the parameter of the measuring points in the vicinity. A statistical magnitude of the parameter of the measuring points in the vicinity is, for example the mean value of the parameter of measuring points in the vicinity. The number of measuring points present in the vicinity is determined by the Gaussian distribution of the white noise and the accepted risk of taking an incorrect decision. An incorrect decision is to be understood to mean a decision whereby the value of the measuring point to be corrected is unduly replaced by the corrected value. A statistical magnitude of the parameter is, for example the mean value of the values of the parameter of the measuring points in the vicinity of the image point to be corrected. On the basis of these choices, the risk P that the ratio K determined exceeds the predetermined reference K
0
is given by the formula:
P
(
K
>
K
0
)
=
(
1
+
K
0
M
)
-
M
.
(
1
)
in which M is the number of measuring points in the vicinity of the measuring point.
Another possibility consists in using, for example the maximum of the measuring points of the vicinity for the statistical magnitude.
A further version of the method according to the invention is characterized in that the measuring points in said vicinity are chosen so as to be symmetrically situated relative to the measuring point to be corrected. An advantage of the choice of a symmetrical distribution of the measuring points in the vicinity of the measuring point consists in that a statistical deviation in the statistical magnitude of the parameter of the measuring points in the vicinity is thus counteracted.
A further version of the invention is characterized in that the measuring points in said vicinity are chosen in such a manner that the measuring points correspond to positions in a dimension of a k space. The k
x
dimension is a suitable choice for this dimension in the k space. As a result of this choice, the method can be simply implemented in an electronic circuit or in a computer program.
A further version of the invention is characterized in that the corrected value of the measuring point to be corrected is determined by the value of an equivalent measuring point of a MR signal generated by a same an MR imaging pulse sequence. When an equivalent measuring point is not available the value of a measuring point to be corrected is determined again by a new measurement of an MR signal which is generated by repeating a similar MR imaging pulse sequence.
A further version of the invention is characterized in that the predetermined reference is chosen in relation to a distance from the origin of a position in the k space which corresponds to th
De Haan Joost M.
Den Boef Johannes H.
Fuderer Miha
Ham Cornelis L. G.
Oda Christine
U.S. Philips Corporation
Vargas Dixomara
Vodopia John F.
LandOfFree
Correction of magnetic resonance signals in a magnetic... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Correction of magnetic resonance signals in a magnetic..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Correction of magnetic resonance signals in a magnetic... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2518132