WO2000010025A1

WO2000010025A1 - Mri image-forming

Info

Publication number: WO2000010025A1
Application number: PCT/BE1999/000109
Authority: WO
Inventors: Hilde Bosmans; Guy Marchal
Original assignee: Katholieke Universiteit Leuven
Current assignee: Katholieke Universiteit Leuven
Priority date: 1998-08-14
Filing date: 1999-08-16
Publication date: 2000-02-24
Anticipated expiration: 2001-02-14
Also published as: AU5402999A

Abstract

The invention relates to image-forming based on magnetic resonance (MRI) of a concealed object, wherein the object and its surrounding area display different relaxation properties, comprising of: applying a magnetic field, aligning magnetic spins of the object and the surrounding area; measuring T1 relaxations by disrupting the alignment with RF pulses and detecting relaxations corresponding with alignment; measuring T2 relaxations by disrupting the alignment with RF pulses and detecting relaxations transversely of the alignment; forming on the basis of the different relaxation properties of the object and its surrounding area at least one image part of the slice from each of the measurements; and performing both measurements for each image part in a predetermined sequence of the detection for a preceding respectively following image part.

Description

MRI IMAGE-FORMING

The present invention relates to a method for forming, on the basis of MRI, images of slices of at least one concealed object, such as an organ in a thorax. The object and its surrounding area display separate magnetic relaxation properties which can be measured by applying a magnetic field and herein aligning spins of magnetic components of the object and its surrounding area and by disrupting the alignment, wherein the degree of relaxation after a determined period of time has elapsed is characteristic for the object or its surrounding area. Use is made for the disruption of RF pulses and the relaxation can be measured in two directions, i.e. parallel to the alignment and transversely of the alignment. In the majority of image-forming techniques (MRI) based on magnetic resonance, image contrast is based in the known art on the TI and T2 relaxation times of the object and its surrounding area, in particular separate tissues. The applying of a strong static magnetic field is of essential importance for this purpose. The microscopic magnetic components or spins in the object or its surrounding area, in particular tissues, are aligned with this field. During an MRI examination this alignment is disrupted by means of resonant RF pulses and the relaxation time of the object and its surrounding area, in particular tissues, is measured towards the equilibrium situation. The TI relaxation time is defined as the recovery of the longitudinal component of the magnetization in accordance with the applied static magnetic field in the object and its surrounding area, in particular tissues, after an RF pulse. The T2 relaxation time relates to the disappearance of the transverse component in the plane transversely of the static field. In practice the object and its surrounding area are adjacent tissues, wherein the separate tissues are characterized by separate TI and T2 relaxation times.

These two parameters are not coupled. The contrast can be increased by obtaining images on the basis of both TI and T2 relaxation times. These two images are then studied together after having been obtained separately. By combined use of TI and T2 relaxation times images can be obtained with which detection and characterization of the object and its surrounding area, usually tissues, are possible. Other properties can also be shown, such as perfusion, diffusion, flow, magnetization current and so on.

According to a first known art, magnetization is first tilted into the transverse plane with a 90° RF pulse. At this moment the longitudinal magnetization amounts to 0. Relaxation begins immediately and differs for the object and its surrounding area, usually different tissues. Materials with a short TI time recover quickly, while materials with a longer TI time recover more gradually. After a time duration of about 0.5 sec. there is optimal contrast between different materials, particularly when they are tissues, and this is therefore the optimal point in time for data acquisition.

In the practice of MRI a sequence of measurements is required for data acquisition. With a single pulse the measurement or measurements can thus be performed on the TI or T2 relaxation for an image pixel, an image line or a whole image.

According to the known art measurements are first performed to obtain measurement values of the TI or T2 relaxation for all pixels, image lines or whole images, only then followed by measurements to obtain measurement values of the other relaxation time for the same (for a correct image-forming) pixels, image lines and whole images . The known art has the problem that pixels, image lines and whole images of the second measurement can be located at a distance from those in the first measurement, this being known as misregistration. This may be the result of breathing, heartbeat, involuntary movement by the patient during a prolonged examination etc., in that there is sufficient time between the measurements for this to occur. Formulation of a radiological protocol based on the combined information of TI and T2 weighted images is hereby made more difficult. Furthermore, as a result of this misregistration, in the case of subsequent superimposing of the images based on Tl and T2 relaxation, the contrast may on the contrary be worsened hereby and even incorrect, undesired and perhaps irrelevant artefacts may herein be displayed in the combined image.

When a formed image is incorrect or inadequate, the examination has to be performed again. This results in a reduction in the maximum capacity of the number of examinations per time unit, whereby the costs of the examination are high. It is moreover the case that if the patient is being examined internally, this patient will feel very uncomfortable as a result of being positioned in the interior of the large and overwhelming device used for the examination.

According to the invention all problems associated with misregistration, as well as misregistration itself, are resolved in a method which is defined according to the present invention in claim 1. Image part herein represents image pixel, image line or whole image depending on the acquisition technique used. In the method according to the present invention the time duration between acquisition of measurement data for each image part and the associated chance of misregistration is minimized, so that the average examination time and the discomfort of a patient for examining are also reduced. The reliability of the superimposed formed image and/or the simultaneous analysis of the Tl and T2 weighted (separate) images is moreover increased.

According to the present invention use is preferably made of techniques wherein the measurement data for the whole image is obtained on the basis of a single Rf excitation pulse, which techniques are known as "single shot" techniques. The duration of the examination is hereby further decreased and the capacity further increased.

In a first preferred embodiment the method according to the present invention is such that measurement of the T2 relaxations precedes measurement of the Tl relaxations and the time between the two measurements is sufficiently long to measure the T2 relaxations, while the alignment-disrupting RF pulse for measuring the T2 relaxations forms a preparation pulse for measuring the Tl relaxations. The phenomena associated with the Tl relaxation as well as T2 relaxation are herein initiated with a single RF pulse. According to this embodiment of the invention optimal use is made hereof in that the phenomena of Tl relaxation can freely develop further, while the phenomena of T2 relaxation are measured at or around a point in time for instance 60-100 msec, after the RF pulse. After a period of time, for instance 0.5 sec. the Tl relaxation phenomena have progressed so far that differences between separate materials or in particular tissues can be well detected. This means that a high contrast in the Tl image can be obtained. A direct measurement of the degree of Tl relaxation is hampered by the applied static magnetic field. In the present preferred embodiment of the method according to the present invention a specific RF pulse for measuring the Tl relaxation can therefore take place transversely of the applied static magnetic field in order to have the measurement of the Tl relaxation take place as shortly as possible after the occurrence of this RF pulse. A very compact and robust method is thus provided. Use can moreover be made herein of a single detector for both the Tl relaxation and the T2 relaxation, since the Tl relaxation is also measured in the direction transversely of the alignment.

The problem of misregistration of slices or image parts can be resolved per se in some applications with so-called navigation pulses. An MRI acquisition is herein divided into different parts, where measurement values are obtained when the position of a diaphragm is situated within a determined range. This procedure can be performed separately for the measurements of the Tl relaxation and the T2 relaxation and the acquisition of images therefrom. The technique is however not robust and the whole procedure is time-consuming. This solution is particularly unsuitable for (relatively minor) injuries outside the thorax or chest, for instance in the abdomen, and other similar phenomena. A description formulated with reference to the annexed drawings is given hereinbelow of a preferred embodiment of a method according to the present invention. In the drawing: fig. la and lb show image-forming results according to the known art and fig. lc and Id show image-forming results according to the present invention; fig. 2 shows a model of spin vectors which is used to describe the effect of resonant RF pulses and the relaxation occurring after such an RF pulse; fig. 3 shows a measurement scheme for acquisition of data of the Tl relaxation; fig. 4 shows a measurement scheme for acquisition of measurement data of the T2 relaxation; fig. 5 shows a measurement scheme in a preferred embodiment of the present invention; fig. 6 shows a flow chart of a further embodiment of the present invention; fig. 7 shows a schematic representation of a test arrangement; fig. 8a-8f show the image-forming results in different acquisition techniques; fig. 9 show the results of image-forming in the case of a living test subject.

Fig. lc and Id show the image-forming results in a method according to claim 2, wherein on the basis of one RF excitation pulse measurement results of both Tl relaxation and T2 relaxation are obtained from the same slice, whereafter separate images are formed on the basis of respectively the Tl relaxation and the T2 relaxation.

The method begins with ultrafast acquisition of measurement data of the T2 relaxation, such as with a single RF pulse "rapid acquisition relaxation enhancement RARE" measurement, "half Fourier acquisition of single shot turbo spin echoes or haste" acquisition or "echo planar imaging" (EPI). These measurements begin with a pulse having an orientation of 90° relative to the static magnetic field and the measurement is completed in a total time for acquisition of about 0.5 sec. With the initial RF pulse for the measurements of the T2 relaxation the longitudinal magnetization is made 0, as shown in fig. 2. The process of the Tl relaxation herein starts immediately, as does the T2 relaxation, and the Tl relaxation has no further influence on the measurements of the T2 relaxation. As shown in fig. 4, image-forming visualizes the different relaxation properties of different tissues, wherein the degree of relaxation after a determined time has elapsed is used to distinguish between separate tissues. The same applies for the Tl relaxation, as shown in fig. 3. In the measurement of the Tl relaxation a second RF pulse is used for the purpose of the measurement in order to make the degree of Tl relaxation measurable in the direction transversely of the static magnetic field, where the magnetic field would form an obstacle to the measurement without such a second RF pulse. Measurement of the Tl relaxation takes place immediately after the second RF pulse when differences in the as it were tilted TI relaxation represent most reliably the different degree of TI relaxation for the separate tissues immediately prior to this second pulse.

In particular embodiments of the present invention use can be made of RF pulses which invert the prevailing magnetization during the T2 weighted measurement. The Tl relaxation could however be delayed herein. This can be remedied in simple manner by observing a short time delay TD, for instance shorter than one second, between the T2 and the Tl weighted measurements, as shown in fig. 5 and fig. 6.

Very fast acquisition methods can also be applied for the measurement of Tl relaxation. In practice the acquisition of the measurement results of the T2 relaxation and of the Tl relaxation for forming separate images or a combined image can therefore be completed in less than 1 second, because "single shot" techniques are used for both types of measurement results. Stated more generally, the T2 weighted measurement can be followed by a Tl weighted acquisition scheme optionally built up of different parts, so-called segmented k-space acquisitions .

The procedure is repeated for different successive slices, as is also the case in other single shot MRI techniques . Of importance herein is that the acquisition of measurement data for T2 relaxation is followed by acquisition of measurement data of Tl relaxation. This can in principle also take place in reverse order according to the invention, wherein it should be noted that the advantage of the sequence followed in this embodiment is that the RF pulse for acquisition of measurement data of T2 relaxation is also the excitation pulse with which the Tl relaxation is initiated, measurement data of which is obtained after acquisition of the measurement data of the T2 relaxation has finished or been completed. By causing acquisition of both measurement data to take place in immediate succession a considerable time-saving and an associated improvement in the quality and reduced danger of misregistration is already achieved, although by following the sequence of the present embodiment these advantages are enhanced even further.

The method was tested by making use of a Haste acquisition with an effective echo time TE of 60 msec, (centre of measurement of the T2 relaxation) as the ultrafast measurement of the T2 relaxation, which was followed after a time of 350 msec, at a delay of 150 msec, by a Haste measurement with an echo time TE = 29 msec. The image obtained with this latter Haste measurement is a measurement of the degree of Tl relaxation as a result of the specific preparation of the longitudinal magnetization by the RF pulse for excitation of the T2 relaxation and of the use of the short TE.

The contrast in the image was verified on the basis of a test arrangement made up of tubes with separate concentrations of contrast agent, which thus exhibit different relaxation times. Fig. la and fig. lb show respectively the images from this test arrangement obtained on the basis of measured Tl relaxation and T2 relaxation, wherein use was made of a known acquisition scheme. The image of the Tl measurement shows that the tube with plain water generates a very hypo-intense signal. This signal increases with an increasing concentration of contrast agent until the concentrations of contrast agent become so high that shortening of the T2 relaxation begins to impair this signal. The image on the basis of measured T2 relaxation shows that the tube with plain water generates hyper-intense signals. For shorter T2 relaxation times the signals decrease gradually.

The test arrangement is shown in fig. 7, in which the concentrations of contrast agent are indicated. The same properties are to be found in the images obtained with the new acquisition scheme. The image on the basis of measured Tl relaxations is shown in fig. lc and compares very well to fig. la. The image of fig. Id obtained on the basis of T2 relaxation is comparable to fig. lb. It will be apparent that the signals from the tube with plain water are even more hyperintense as a consequence of an effective repetition time which equals infinity.

Starting from the desired application and the used algorithms, the optimal point in time at which the acquisition of measurement data of the Tl relaxation should begin can be determined.

Fig. 5 shows schematically together with fig. 6 the method in which use is made of the acquisition, wherein two measurements of the T2 relaxation weighted information take place in the form of an "early and late TE haste" algorithm. After a single or, as shown here, double acquisition of measurement results for T2 relaxation, collecting of measurement data takes place for the Tl relaxation after a time TD which is sufficient to enable imaging of a good contrast for the different tissues in the measurement results for the Tl relaxation.

Methods of obtaining measurement results are defined on the basis of the time rhythm of RF pulses, the application of magnetic field gradients (the strength of the gradients and the rise times) and the operation of the detector. The operation of a specific scanner, such as a 1.5 T MR scanner (vision, Siemens, Germany) can be controlled by means of special software packages, for instance the software package called "pargen" . Fig. 6 shows a flow chart of the structure of the method also schematically shown in fig. 5, wherein double measurement of the T2 relaxation phenomena takes place. The first Haste acquisition for measurement values of the T2 relaxation begins with an RF excitation pulse with an orientation which differs 90° from that of the static magnetic field and results in a series of signal read-outs alternately having different phase encodings, as required for the Fourier transformation based image reconstruction. For an image with M rows the k-space is filled with M/2 + 8 lines, starting from line -8 up to M/2 and is thus incompletely filled. The time between successive signal read-outs is 3.6 msec. The acquisition process begins after a time duration of 30 msec. The effective echo time (TE) , i.e. the time between the RF excitation pulse and scanning of the central k-space line, is 60 msec. For an image matrix with 160 x 256 pixels, 88 signals are measured. This requires an acquisition window of 88 x 3.6 ms = 316 ms . The measurement is continued, filling a second k-space, again from line -8 up to M/2. The total acquisition window for this image again amounts to 316 ms . A total time of 663 ms has thus elapsed since the occurrence of the RF pulse. In the acquisition, as shown in fig. 5, 180° RF pulses are used to obtain T2 weighted spin echo images. This is advantageous because the two Haste measurements for the T2 relaxation lasted longer than the time duration optimal for acquisition of the Tl relaxation of 0.5 s for the most common tissues (see above) . No information is therefore lost relating to the optimal contrast .

Such 180° RF pulses are also used during acquisition of measurement values of the Tl relaxation. The time duration available for acquisition of measurement values of the Tl relaxation is hereby increased because the Tl relaxation is here also delayed.

For the purpose of measuring the longitudinal magnetization, or the relaxation in the direction of alignment, a new RF pulse with an orientation differing 90° from the alignment is used to transfer the longitudinal magnetization prevailing at the moment of the pulse into the transverse plane transversely of the alignment. The static magnetic field thus forms no obstacle in detection of the longitudinal magnetization, since it is transferred in the transverse plane. A third k-space is then filled with data which is phase encoded and the signals are read out in a manner such that the k-space between line -8 and line M/2 is filled. The time duration between read-out of successive signals also amounts to 3.6 msec, wherein the effective echo time TE 3 is 29 msec, which is defined as above as the time from the RF pulse to the scanning of the central line of the k-space. A test was performed making use of the same test arrangement as that on which fig. la to fig. Id is based, i.e. that shown in fig. 7. In fig. 7 the ratios indicate solutions of Gd-DOTA. This is a contrast agent. Fig. 8a, b and c show images formed on the basis of measurement values obtained with known acquisition techniques. For fig. 8a, which was formed on the basis of the acquisition of measurement results of Tl relaxation, TR amounted to 500 msec and TE to 30 msec Fig. 8b shows the image formed on the basis of measurement results of T2 relaxation with TR = 2000 msec, and TE = 60 msec, while fig. 8c is a late acquisition of measurement results of T2 relaxation with TR = 2000 msec and TE = 120 msec.

Fig. 8a shows that the tube with plain water generates very hypo-intense signals as a result of the exceptionally long Tl relaxation hereof. This signal increases with an increase in the concentration of the contrast agent until the doses are so high that shortening of the T2 relaxation time has a detrimental effect on this signal. The acquisition of the measurement results for T2 relaxation at TE = 60 msec. shows that the tube with the pure physiological solution has hyper-intense signals. For shorter T2 relaxation times with increasing concentrations, the signals gradually decrease, as shown in fig. 8b. The same is found in fig. 8c as the result of the image-forming on the basis of acquisition of measurement results for the T2 relaxation at a late echo time (TE) .

The same properties are found in the images obtained with the acquisition method according to the present invention. The image obtained on the basis of measurement results of the Tl relaxation is shown in fig. 8d and the image contrast is readily comparable to that of fig. 8a. Fig. 8e is comparable to fig. 8b. It is found that the signal from the tube with plain water is even more hyper-intense as a consequence of an effective repetition time, which becomes equal to infinity.

Finally, the image on the basis of measurement results of the T2 relaxation at a late TE (echo time) in fig. 8f can be compared with the known acquisition, wherein use is made of an echo time (TE) of 120 msec. , the image of which is shown in fig. 8c

Fig. 9 shows the results of the in vivo test of the acquisition of measurement results of the Tl relaxation, of the early acquisition of measurement results of the T2 relaxation and of the late acquisition (with a high TE) of measurement results of the T2 relaxation, which were obtained from a normal volunteer. The image-forming corresponds with the images usually obtained with other separate measurements of Tl relaxation, T2 relaxation, and late T2 relaxation. The figures moreover show clearly that the three formed images are obtained from exactly the same anatomical position. As a result of the method according to the present invention there has occurred no shift at all in this perspective which could result in misregistration or shift or movement errors. Breathing and other involuntary movements therefore have no or only very little effect thereon.

The present invention as defined in the appended claims is not limited to the above described embodiments. Use can thus be made of a separate RF pulse for acquisition of measurement values of the Tl relaxation, while in the above described embodiments the RF pulse was used each time for excitation of the T2 relaxation. When 180° pulses are used, as in spin echo measurements, a slight delay is then caused between both acquisitions, which can however be kept small. Methods other than those based on a single pulse for obtaining an entire image can also be used, for instance those where with each pulse the data for one image line is collected. Of importance here according to the present invention is that acquisition of measurement values of Tl relaxation and of T2 relaxation are performed per image line, wherein both acquisitions are performed prior to and/or subsequent to those performed for another image line.

Claims

1. Method of image-forming based on magnetic resonance (MRI) of slices of at least one concealed object, such as an organ in a thorax, wherein the object and its surrounding area display different relaxation properties, comprising of:

- applying a magnetic field and hereby aligning spins of magnetic components of the object and its surrounding area;

- measuring Tl relaxations by disrupting the alignment with at least one RF pulse and detecting relaxations of the spins occurring in the direction of alignment;

- measuring T2 relaxations by disrupting the alignment with at least one RF pulse and detecting relaxations of the spins occurring in the direction transversely of the alignment;

- forming on the basis of the different relaxation properties of the object and its surrounding area at least one image part or image of the slice from each of the measurements of the Tl relaxations and/or the T2 relaxations; and

- performing both measurements for each image part in a predetermined sequence subsequent to and/or prior to the detections for a preceding respectively following image part or image.

2. Method as claimed in claim 1, wherein measurement of the T2 relaxations precedes measurement of the Tl relaxations and the time between the two measurements is sufficiently long to measure the T2 relaxations, while the alignment-disrupting RF pulse for measuring the T2 relaxations forms a preparation pulse for measuring the Tl relaxations.

3. Method as claimed in claim 2, further comprising of measuring the Tl relaxations after the RF pulse for measuring T2 relaxations.

4. Method as claimed in claim 1, 2 or 3, wherein the RF pulse for measuring at least the Tl relaxations causes a disruption of 90° relative to the alignment.

5. Method as claimed in claim 4, further comprising of measuring the Tl relaxations in a direction transversely of the alignment.

6. Method as claimed in any of the foregoing claims, further comprising of measuring Tl relaxations or T2 relaxations, on the basis of which a single image can be generated after a single alignment-disrupting RF pulse.