FI100036B - Dynamic magnetic resonance imaging method - Google Patents

Description

100036

Dynamic magnetic resonance imaging method. - Dynamiskt magnetav- bildningsförfarande.

The invention relates to a magnetic resonance imaging method which is particularly well suited to e.g. to describe dynamic phenomena in the human body. An example of such a dynamic phenomenon is the distribution of contrast in the body after injection.

10 Magnetic resonance imaging has been discussed in several publications. Reference is made here to Raimo Sepponen: Imaging method, US-5,227,723.

The dynamics of the distribution of the contrast medium has been used in the diagnos-15 splash e.g. to distinguish between benign and malignant busts. Experiments have shown that the contrast change of a malignant tumor reaches more than 90% of its maximum in less than one minute after injection. This requires a fast T1-weighted shooting method. It is known to use in this case a so-called "keyhole" technique, in which low-resolution images are successively described using signals with relatively little phase coding. The problem with this technique is not only the poor resolution of the resulting image but also the artifacts, which are strong in the phase coding direction. In addition to movement, so-called truncation artifacts are also strong.

The effective time difference of successive images captured by the Keyhole technique is of the order of the imaging sequence. In this case, it can be several tens of seconds. Thus, a series of images does not form a continuous representation of the dynamic event itself. In other words, the time resolution of the sequence is poor.

With the method according to the invention, the disadvantages of the prior art 35 can be avoided and a series of magnetic images of a dynamic phenomenon can be obtained so that the effective time difference between successive images is only 100036 2 (tens of milliseconds) of sequence repetition time, i.e. the time resolution of the sequence is good. In addition, the spatial resolution of the achievable images is good.

The invention is characterized by what is apparent from the characterizing parts of the claims.

The invention is illustrated by the accompanying drawings, in which 10 - Figure 1 shows a pulse sequence according to the invention - Figure 2 shows a signal acquisition strategy of a pulse sequence according to the invention in k-space.

Figure 1 shows a pulse-15 diagram of the imaging method according to the invention. The tuning pulses are described by the RF axis. The time between them is the repeat interval TR. The signals emitted by the nuclei of the target are represented by the axis S. The echo signal is SE and can be generated either by the so-called gradient inversion or by the pulse technique. In the pulse sequence shown in Figure 1, the signal 20 is collected as an echo signal generated by gradient inversion. If it is desired to generate a pulse echo, a so-called refocusing pulse must be applied to the target, which is usually a so-called 180 ° pulse.

These issues are discussed in the aforementioned U.S. Pat. No. 5, 227, 723 and in several references cited therein.

25 In principle, the echo signal may not need to be used or the echo may be asymmetric. The signals are collected such that λ J that the direction of the gradient during signal acquisition is during different signal acquisition events.

Figure 2 shows the course of signal acquisition trajectories in k-space. Each signal passes through the center of the k-space. In this case, it can be thought that the Fourier transform of each signal is the projection of the object with respect to the direction of the gradient. The image of the object can be reconstructed by collecting several 35 projections and using the so-called projection reconstruction technique in calculating the image. A low-resolution image can be: calculated with 32 projections. If TR is 100 ms then:: sd; i ti · t t-ι rt 3 The first image of 100036 is obtained about 3 seconds after the start of recording. The next image can be formed after as little as 100 ms using the projection just collected and possibly omitting the first projection. Thus, by continuing the 5-la, images with an average time difference of 100 ms or even shorter can be collected, but more accurate images can also be calculated from the projections, e.g., using 128 projections. The first such image is obtained after 13 seconds and the following after every 100 ms. Finally, all projections can be used to produce an accurate anatomical image with a good signal-to-noise ratio. The set of signals that are thus collected makes it possible to form different image sequences: Optionally, in the image sequences, the spatial resolution of the individual images may be low but the time resolution of the series 15 may be good. A preferred embodiment is to form an image of the collected signals with good position resolution and signal-to-noise ratio, and to display lower resolution images superimposed on the image with good position resolution as a continuous sequence of images.

20

Figure 2 illustrates the method of signal acquisition according to the invention by showing the course of the trajectories (projections) P of the signal set acquisition. Projection indices 1 to N refer to the projection angle associated with each of the signals 2 to 5. The letter index indicates the trajectories belonging to different signal subsets. Thus, a first image is formed of the signals collected according to the trajectories P1, P2 ... PN, a second image of the signals of the trajectories Pia (= P2), P2a (= P3) .... and a third image of the signals of the trajectories P1b (= P2a = P3), etc. Thus, a set of signals is collected from the object, from the subsets of which the image set is reconstructed. Said signal subsets are not identical and the signals of the subsets are selected so that the averages of their collection times differ from each other. In this case, the images formed from the subsets 35 represent the object on average at different times. According to the invention, an object is formed: an image with good resolution and signal-to-noise ratio, 4 100036 using all or at least a significant part of the collected signals. Figure 2 shows only one way of collecting signals, it may be advantageous to divide the k-space in different ways in order to achieve e.g. better time resolution.

5

With the method according to the invention, the dynamics of the distribution of the contrast medium can be studied. Studies have found e.g. a malignancy in the breast rapidly takes up the contrast agent. In this case, more than 80% of the maximum enhancement is reached within the first minute of the contrast agent dose. With benign tumors, intensification is slower. A paramagnetic compound (Gd-DTPA, Magnevist (R), Schering AG,

Germany).

15

Images can be generated, for example, using a projector construction technique or by interpolating a rectangular matrix from the measurement results and reconstructing the image using a Fourier transform.

20

The trajectories of the signals collected according to the invention always pass through the origin of the k-space. This achieves maximum sensitivity to contrast changes. The weakness of the so-called keyhole method according to the prior art is that the trajectories of the -25 signals do not pass through all but one of the centers of the k-space. In this case, the sensitivity of the images to contrast changes is not at a maximum, and it is not possible to form an image set from the signal set that can continuously monitor the contrast changes.

30

Only one embodiment of the invention has been described above, several other alternative embodiments can be considered within the scope of the inventive idea expressed by the claims.

35 ti: Htf.r itH I M Ί

Claims (4)

100036 A magnetic resonance imaging method for examining an object, such as a human body or a tree trunk, characterized in that it collects a set of signals from the object using a magnetic field gradient during a signal acquisition event. used to form a set of magnetic images. A method according to claim 1, characterized in that a part of said signal subset is formed from the signals of the signal set such that the time interval of the signal acquisition events of the signal part sets is between different different signal subset. A method according to claim 1 or 2, characterized in that a contrast agent such as a paramagnetic agent is applied to the target before the signal set is collected. Method according to one of the preceding claims, characterized in that the images formed from the signal subset are displayed in such a way that at least two images of the magnetic image set are superimposed. 100036