JPH0449418B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION In measuring the nuclear magnetic distribution within a region of an object, the present invention generates a uniform static magnetic field, places the region of the object within this static magnetic field, and measures a) high-frequency electromagnetic b) generating a pulse to precess the magnetization of the nuclei in the object, thereby generating a resonance signal; b) subsequently, after a preparatory period, during a measurement period that is divided into several sampling periods, detecting this resonance; periodically take some signal samples from the signal (FID signal); c) then repeat steps a) and b) a number of times (n′), each time after a waiting period, to collect a group of signal samples; The present invention relates to a nuclear magnetic distribution measurement method for obtaining a distribution image of nuclear magnetization induced therefrom.

The present invention also provides a means for generating a uniform static magnetic field, b) a means for generating high-frequency electromagnetic radiation, c) a means for generating a gradient magnetic field, and d) for determining the nuclear magnetic distribution within a region of an object. sampling means for sampling the resonance signal generated by the means specified in a) and b); e processing means for processing the signal provided by said sampling means to form a nuclear magnetic distribution; f at least the term b; controlling the means specified in ) to e) to generate some resonance signals;
Control means for conditioning, sampling and processing; and a nuclear magnetic measurement apparatus.

A similar method (also called Fourier zeugmatography) and device are known from US Pat. No. 4,070,611. According to such a method, the direction of the field on the object to be inspected is, for example, z in the Cartesian coordinate system (x, y, z).
Apply a strong uniform static magnetic field B ₀ aligned with the axis. This causes a slight polarization of the nuclear spins existing in the object of this static magnetic field B ₀ , and the nuclear spins become polarized by the static magnetic field B ₀
Allows for precession around the direction of.
When the static magnetic field B ₀ is applied and then set at 90°, a suitable pulse of high-frequency electromagnetic radiation is generated (angular frequency ω =
γ·B ₀ , where γ is the gyromagnetic ratio and B ₀ is the magnetic field strength). This rotates the magnetization direction of the nuclei within the object by 90 degrees. After the end of the 90° pulse,
The nuclear spins begin to precess around the direction of the static magnetic field B ₀ and generate a resonance signal (FID signal). Using gradient magnetic fields G _x , G _y , G _z whose field directions coincide with the direction of the magnetic field B ₀ , the total magnetic field is B = B ₀ +
G _x・x + G _y・y + G _z・z, and the strength depends on the location. This is because the strengths of the gradient magnetic fields G _x , G _y , and G _z are gradients in the x, y, and z directions, respectively.

After the 90° pulse, a gradient magnetic field G _x is applied for a period t _x ,
Next, a gradient magnetic field G _y is applied for a period t _y . The precession of the excited nuclear spins is therefore influenced in a location-dependent manner. This preparation period (i.e. t _x + t _y )
After that, a gradient magnetic field G _z is applied and the measurement period t _z is followed by l _xn times.
The FID signal (which is essentially the sum of all nuclear magnetizations) is measured at N _z measurement moments. At that time, the value of t _x and/or t _y is made different each time. In this way, (N _z ×m × l) signal samples are obtained, which contain information about the nuclear magnetic distribution of a region of the object in x, y, z space. l × m sets of N _z signal samples are stored in memory (N _z × m
xl) memory locations, and then performs 3-D Fourier transform on the sample values of the FID signal to obtain an image of the nuclear magnetic distribution.

Obviously, selective excitation could alternatively be used to generate FID signals of nuclear spins only within a two-dimensional slice (direction freely selectable). At this time, for example, it is sufficient to generate the FID signal m times in order to obtain an image of the nuclear magnetic distribution at m×N _z points on the selected cross section by two-dimensional Fourier transformation. As is clear from the above, when using the Fourier chromatography method, it takes at least several minutes to form an image of the nuclear magnetic distribution. Such a long measurement period is unacceptable to the patient undergoing the test, who must remain still during that period.

It is an object of the present invention to provide a method and a device whose resolution is at least equal to that obtainable in the state of the art, for example with Fourier chromatography, yet which considerably reduces the time required to form an image of the nuclear magnetic distribution. It is on offer.

Another object of the present invention is to provide a method and apparatus capable of forming a nuclear magnetic distribution image with a clear intensity distribution compared to an image formed by a pulse sequence used in the current state of the art.

To achieve this objective, the method of the present invention generates a high-frequency 180° pulse after extraction of the sampling signal to generate a nuclear spin echo signal to cancel the influence (phase shift) imposed on the nuclear magnetization by the gradient magnetic field.
It is then characterized in that another high-frequency excitation pulse is generated such that a nuclear spin echo signal is present.

As a result of using another radio-frequency excitation pulse, the measurement cycles can continue more quickly from one to the other without significantly adversely affecting the signal strength of the nuclear spin resonance signal in successive measurement cycles. Although there will be a slight loss of insignificant signal, the generated resonance signal will contain other information leading to a sharper image intensity distribution. As a result, the use of separate high-frequency excitation pulses not only shortens the measurement cycle, but also produces additional image information that depends on the elapsed time between successive pulses.

This additional high-frequency excitation pulse is particularly sensitive to the effects of phase shifts due to field inhomogeneities or other carefully considered inhomogeneities of the uniform static magnetic field (B ₀ field) and of the excited gradient field. Gives the desired result when canceled.

In this method, the waiting time between another high-frequency excitation pulse and the start of the next measurement cycle is approximately as large as the time that elapses between a high-frequency 90° excitation pulse and another high-frequency excitation pulse. This is three times the total time required to take the several samples needed to form an N-M-R image (without significant signal loss) as required by the current state of the art. No
That means it's 10 times shorter.

With the method of the invention, a first gradient field is applied after selectively exciting the fault (by known techniques) during the preparation period, and during the measurement period the direction of the gradient or the direction of the gradient of the first gradient field is The nuclear spin density distribution within the object's cross section is measured by applying different gradient magnetic fields extending perpendicular to the object. Furthermore, in contrast to known methods for determining the nuclear spin density distribution, the method of the present invention allows for position-dependent nuclear spin resonance spectra (position-dependent nuclear spin spectroscopy)
You can also ask for For example, by using selective excitation and applying two gradient magnetic fields (with two orthogonally oriented gradient directions in the excitation plane), the present invention can reduce the nuclear spin resonance per pixel in the slice. Allows to obtain frequency spectra (chemical shifts). The number of pixels in the section can be selected according to the state of the art.

As already mentioned, the precession frequency of nuclear magnetism M is ω
= Determined by γ・B. The gyromagnetic ratio γ depends only on the type of nucleus, as long as the nucleus is considered "free." However, atomic nuclei are not usually considered free. This is because the nucleus is influenced by the bonding electrons around itself. This becomes apparent as a so-called chemical shift σ. The bound nucleus does not resonate at ω=γ・B, and ω′=γ・B×(1
−σ). Nuclear frequency ω′ and detuning Δω=ω
-ω'=γ・B・σ is proportional to the magnetic field B. In general, σ is very small (on the order of magnitude of 10 ^-6 to 10 ^-4 ). When the strength of the magnetic field B is sufficient, for example, 1T or more, the value of σ is different, so
Individual peaks can be identified in the generated spectrum. Each of these peaks is associated with a nucleus having a different chemical bond. For example, in the spectrum of phosphorus (31P) in living organisms, phosphocreatine,
Peaks associated with ATP and inorganically bound phosphorus can be identified (e.g. RLNunnally's "Localized measurement of metabolism by
"NMR methods", The Proceedings of an
International Symposium on NMR (Wake Forest University Bowman)
Gray Medical School, Winston-Salem, North
Carolina, October 1-3, 1981) No. 181-184
(See page listing). The correlation of the magnitudes of these peaks contains information regarding the metabolic state of the tissue. It has proven useful to plot such spectra as a function of position within the tissue or part of the object under examination. For this purpose, the method and apparatus disclosed in the aforementioned US Pat. No. 4,070,611 can be used when the gradient field is not activated during the measurement period (and thus when sampling the resonance signal).

The above will be explained based on an example in which the spectrum is obtained as a function of the coordinates (eg, x, y) of two positions. In the first step, a fault with thickness Δz in the direction perpendicular to the z-axis is excited by selective excitation. Then, during a preparatory period, x and y
or activating either one of the gradient magnetic fields.
Then, the resonance signal occurring during the measurement period is sampled.

As a result, the gradient magnetic field is not activated during the measurement period, and the resonance signal to be measured is at the image frequency K _x ,
It is a function of K _y and time t. As mentioned later,
The image frequencies K _x and K _y are determined by the gradient magnetic fields G _x and G _y applied during the preparation period. By repeating the measurement cycle many times, activating the gradient fields G _x and G _y during the preparatory period, each time with varying strength and/or duration, a series of signal samples of the resonant signal are generated at each different image frequency as a function of time. is elicited for the pair (K _x , K _y ). In this way, a three-dimensional matrix (K _x , K _y , t) is filled with signal samples. Then, after 3-D Fourier transformation (to t, _{K x ,} and _{K y )} from the 3-dimensional data matrix (K A frequency spectrum is found at the point (x,y) within. The method described above can be very easily extended from a two-dimensional plane to a three-dimensional volume. Three gradient magnetic fields are activated during the preparation period after exciting the predetermined volume. Then (in the absence of gradient fields) during sampling the set of frequencies (K _x ,
K _y , K _z ). The preparatory period (or other period during which the gradient field is activated) by changing the strength of the gradient magnetic field (G _x , G _y , G _z ) each time.
By repeating the measurement many times during the test, a four-dimensional matrix (K _x , K _y , K _z , t) is filled with signal samples. After a 4-D Fourier transform (to t, K _x , K _y and K _z ) each point within said volume (x, y,
A frequency spectrum is obtained at z). As previously mentioned, the subject cannot be shifted or moved during the sequential waiting, preparation and measurement times. However, the method of determining the resonance spectral distribution according to the invention described above requires a considerable amount of time to form an image of the local nuclear spin resonance spectrum (at least with a resolution equal to that obtained by known Fourier chromatography techniques). It can be shortened to This is of course an advantage.

A preferred embodiment of the method according to the invention comprises, after another radio-frequency excitation pulse, completing a similar cycle of radio-frequency pulses with associated gradient fields after a period T;
The gradient magnetic field is completed in at least one pulse interval, characterized in that at least one pulse interval during the first cycle is different from the corresponding pulse interval during the second cycle. This preferred embodiment has therefore been found to provide significantly sharper contrast in the final image formed when sampling signals are taken. The steady-state solution to the Brodscho equation shows that a negative signal that causes this contrast increase can occur in the resonant signal during the second cycle.

Apparatus for carrying out the method of the invention includes preprogrammed computer means for generating and supplying control signals to the means for generating high frequency electromagnetic radiation, the control means comprising preprogrammed computer means for generating and supplying control signals to the means for generating high frequency electromagnetic radiation;
An adjustable pulse train of 180° excitation pulses can be generated such that the period of time that elapses between two recently generated 180° excitation pulses is the same as that between the most recent 180° pulse and the next other 90° excitation pulse. The feature is that the period is twice as long as the period that elapses. This device allows the method of the invention to be carried out easily.
If necessary, it can be adapted to the properties of the subject (for example, when using contrast changes in NMR images).

The invention will be explained in detail with reference to the drawings.

FIG. 1 shows a coil system 10 forming part of a device 15 (FIG. 2) used to determine the nuclear magnetic distribution in the region of a human body 20. FIG. This region has, for example, a thickness Δz and the illustrated coordinate system (x, y,
z) in the xy plane. The y-axis extends upward perpendicular to the plane of the paper. The coil system 10 has a uniform static magnetic field B ₀ whose field direction is parallel to the z-axis, and three coils whose field direction is parallel to the z-axis and whose gradient directions are parallel to the x, y, and z axes, respectively. gradient magnetic fields G _x , G _y and G _z and a high frequency magnetic field. To do this, the coil system 10 includes a set of main coils 1 for generating a static magnetic field.
Equipped with. The main coil 1 is arranged, for example, on a spherical surface 2 whose center is at the origin 0 of the illustrated Cartesian coordinate system x, y, z, and the direction of the main coil 1 coincides with the z-axis.

The coil system 10 also comprises four coils 3a, 3b for generating a gradient magnetic field _Gz . To do this, the first set of coils 3a is replaced by the second set of coils 3b.
It is excited by a current in the opposite direction to the current flowing inside. This is indicated by ○・ and ○× in the figure, but ○ is the coil 3
○× means the current flowing into that part of the coil.

The coil system 10 also comprises four rectangular coils 5 (of which only two are shown) or other four coils, for example "Golay coils", for generating the gradient field G _y . . In order to generate the gradient magnetic field _G ing. FIG. 1 also shows a coil 11 for generating and detecting high frequency electromagnetic fields.

FIG. 2 shows an apparatus 15 for carrying out the method of the invention.
shows. The device 15 includes the coils 1, 3, 5, 7 and 11 described with reference to FIG.
Current generator 17,1 for exciting 5 and 7
9, 21, and 23, and a high frequency signal generator 25 for exciting the coil 11. The device 15 also includes a high frequency signal detector 27, a demodulator 28, a sampling circuit 29, processing means 31 such as an A/D converter, a memory 33, an arithmetic circuit 35 for performing a Fourier transform, and a sampling circuit 29. A control device 37 for controlling the instant, a display device 43,
It also includes a central control means 45 whose functions and relationships will be described later.

The device 15 implements a method for determining the nuclear magnetic distribution within the region of the human body 20 as described below. This method involves periodically repeating a measurement cycle that can itself be divided into several stages. During the measurement cycle a portion of the nuclear spins present in the human body are resonantly excited. This resonant excitation of the nuclear spins is obtained by activating the current generator 17 in the central control unit 45, so that the main coil 1 is energized and remains energized for the desired number of measurement cycles. In this way, a uniform static magnetic field B ₀ is generated.
Also, the high frequency generator 25 is switched on for a short time so that the coil 11 generates a high frequency electromagnetic field. Nuclear spins within the human body are resonantly excited by an applied magnetic field. The excited nuclear magnetism assumes a predetermined angle, for example 90°, with respect to the direction of the uniform magnetic field B ₀ (90° radio frequency pulse). Which nuclear spins are excited depends on the strength of the magnetic field B ₀ and the angular frequency ω ₀ of the gradient magnetic field and high-frequency electromagnetic field to be applied.
Depends on. γ is the gyromagnetic ratio (for free protons, e.g. H ₂ O protons, γ/2π = 42.576MHz/
This is because the equation ω ₀ =γ·B ₀ (1) must be satisfied as T). After the excitation period, the central controller 4
5 switches off the high frequency generator 25. Resonant excitation always takes place at the beginning of each measurement cycle. However, in some operating methods, high frequency pulses are induced within the human body even during the measurement cycle. In this case, these high-frequency pulses are, for example, a pulse train consisting of periodically generated 180° high-frequency pulses. The latter is called a "spin echo." Spin echo is described in an article by ILPykett titled ``NMR in Medicine'' published in the May 1982 issue of the Chinese magazine ``Scientic American.''

In the next step signal samples are collected.
For this purpose, gradient magnetic fields generated by current generators 19, 21 and 23, respectively, under the control of central control unit 45 can be used. A resonance signal (referred to as an FID signal) is detected by a high frequency detector 27, a demodulator 28, a sampling circuit 29, and an A/D converter 31.
and by switching on the control device 37. This FID signal appears as a result of the precession around the field direction of the nuclear magnetic field B ₀ caused by the radio-frequency excitation pulse. This nuclear magnetism induces an induced voltage in the detection coil,
Its amplitude is a measure of the strength of nuclear magnetism.

The analog sampled FID signal originating from sampling circuit 29 is digitized by A/D converter 31 and stored in memory 33. After the last signal sample has been taken at the instant t _e , the central controller 45 connects the current generators 19, 21 and 23 to
The sampling circuit 29, the control device 37, and the A/
The D converter 31 is stopped.

The sampled FID signal is in memory 33 and continues to be stored there. Then, the next measurement cycle is performed, during which the FID signal is again generated, sampled and stored in memory 33. When a sufficient number of FID signals have been measured (the number of FID signals to be measured depends, for example, on the desired resolution), a 2-D or 3-D Fourier transform (either one under the influence of The FID signal is generated and sampled by using gradient magnetic fields) to form an image.

Figure 3a shows an example of a measurement cycle according to the current state of the art. This will also be explained with reference to the device 15 shown in FIG.

After energizing the main coil 1 which generates a uniform static magnetic field B ₀ , the high frequency coil 11 is used to generate a 90° pulse P1. The resonance signal F generated by this
1 is allowed to attenuate when using the spin echo technique, and after a certain time tv1 a 180° pulse P2 is generated by the high frequency coil 11. For reasons to be described later, gradient magnetic fields G _x and G _y shown by curves G1 and G3 are generated during a part of period tv1, respectively.
180° pulse P2 after a period tv2 of length equal to tv1
The echo resonance signal F2 generated by this reaches a peak value. So-called spin echo technology (180° pulse P2)
Using , phase errors in the resonance signal caused by the spin nuclei are avoided. Such a phase error is caused by the non-uniformity of the static magnetic field B ₀ . The echo resonance signal is sampled after each sampling time t _n in the presence of a gradient field G _x as indicated by curve G2.

It is known that the magnetic phase angle at point x in the gradient magnetic field G _x is given by the following equation.

∫ ^t γ・G _x・x・dτ Therefore, the image frequency K _x can be defined as follows.

K _x =γ·∫ ^t G _x ·Dτ Therefore, after each sampling period t _n the respective signal sample associated with the corresponding (different) image frequency K _x is determined. The sequential image frequencies exhibit an image frequency difference ΔK _x =γ·∫ _tn G _x ·dτ. Obviously, the measurement cycle described above is repeated, with the gradient field G _y being applied for some time before sampling, and signal samples relating to the image frequency pair (K _x , K _y ) are obtained.
If there is no gradient magnetic field G _y , the image frequency (K _x ,
0) are obtained. The image frequency range is −K _xn ~ +K _xn and −K _yn ~ +
When a group of signal samples related to a matrix of image frequencies K _{x and} K _y , which is K _yn , is collected, the distribution of nuclear magnetism in the xy plane can be determined from this group of signal samples by 2-D Fourier transformation. Show what you can do. Here |K _xn | and |K _yn
| is the highest image frequency occurring within the matrix. Therefore, in order to determine the distribution of nuclear magnetism, it is necessary to extract signal samples associated with image frequencies between -K _xn and +K _xn and between -K _yn and +K _yn . The image frequency K _y is determined by K _y = γ∫ ^ym _tv1 G ₃ (τ).

It therefore has a constant value during the measurement cycle.
The image frequency K _x is determined by the gradient magnetic fields G1 and G2. The strength and duration of these gradient fields are adapted to each other such that at instant t ₀ a signal sample associated with the image frequency pair (0, K _y ) is extracted. This means that the following formula holds.

∫ _tv1 γ・G ₁ (τ)dτ=∫ _tv2 γ・G ₂ (τ)dτ Integral∫ _tv1 γ・G ₁ (τ)dτ Adjusted so that this integral is equal to +K _n , the instant t=t _s The first signal sample extracted at is the image frequency pair (−K _xn ,
Related to _Kyn ). After the time T of the measurement cycle started with pulse P1 has elapsed, a next measurement cycle is started with a similar measurement pulse P1' and a new sequence of signal samples associated with the image frequency pair (K _x , K _y ) is taken. Now during this measurement cycle K _x is variable, but K _y is constant,
It is predetermined that during the period tv1' between pulses P1' and P2', a gradient field G1' and a gradient field G3' (different in strength during each measurement cycle) are applied. In the state of the art, the time T between the starting points of two measuring cycles amounts to 0.5 to 1 second. Attempts to further reduce this time will come at the cost of nuclear spin echo signals generated during the next measurement cycle. This is because a significant portion of the capped and excited nuclear spins has a relatively long relaxation time with respect to this shortened time. Therefore, only the portion of the nuclear spin relaxed in the direction of the main magnetic field B ₀ contributes to the next spin echo signal.

As shown in Figure 3b (all pulses in Figure 3b that are also shown in Figure 3a are given corresponding symbols), if there is no gradient magnetic field during the measurement period,
Signal samples are taken that are a function of K _x , K _y and time t. Using the measurement cycle shown in FIG _. 3b, one row after the other rows _of the three-dimensional matrix (K ) is derived. This is a frequency spectrum (f) that depends on the (x,y) position.

It should be noted that the time required to carry out the method shown in FIGS. 3a and 3b can be considerably reduced, as will be discussed below with respect to FIGS. 4 and 5.

FIG. 4 shows the measurement cycle of the method of the invention. This measurement cycle essentially consists of Figures 3a and 3b.
This is the same as one of the measurement cycles shown in the figure. Note that for the sake of clarity, FIG. 4 does not show the gradient magnetic field, but only the associated nuclear spin echo signal. However, in the method of the invention, after the echo instant of the generated nuclear spin echo signal F2, a time t
After 2, a high frequency 180° pulse P3 is generated. Nuclear spin echo signal F3 generated by this
Another high frequency excitation pulse P4 is generated during the period. It is preferable that the center of gravity of the high-frequency excitation pulse P4 (in the form of a Gaussian pulse) coincides with the instant of the echo of the echo signal. Pulses P3 and P4 serve the following purpose. If inhomogeneities in the gradient field and other (carefully induced) fields are compensated for, the 180° pulse P3 is in phase (at the instant of echo) with the main field B ₀
Insert the nuclear magnetic component that is oriented perpendicular to the As a result, the direction of these components of the nuclear magnetism can be easily changed by another high-frequency excitation pulse P4. The phase of the high-frequency excitation pulse P4 relative to the composite phase of the nuclear magnetism at the moment of the echo determines the change in direction of the components of the nuclear magnetism that are now in phase. (of selectivity)
Assuming that the phase of the electromagnetic pulse P1 is 0° and defining this pulse P1 as an x pulse, the high frequency 180° pulse P2 is generally an x pulse or -x pulse (in phase or opposite phase to P1), or a y pulse or -
It becomes a y pulse. However, pulses with different phases can also be used.

When the high frequency 180° pulse P2 is an x pulse, the second echo signal F3 is converted to a −x180° pulse P3.
It is preferable to generate this together. In this way, the first
can compensate for the effects of the inhomogeneity of the radio-frequency magnetic field and the static magnetic field _B0 , which adversely affect the strength and phase of the nuclear spin echo signal F2. At this time, the two sequential nuclear spin echo signals F2 and F3 are, for example, −y
phase and +y phase. During the second nuclear spin echo signal F3, another high frequency 90° (selective) excitation pulse P4 is generated, the phase of which determines the directional change that the nuclear magnetism undergoes. Pulse P
If 4 is a −x pulse, the perpendicular component of the nuclear magnetism is rotated to point in the direction (of the positive z-axis) of the static magnetic field B ₀ . For +x pulses, these vertical components are oriented in the opposite direction (negative z-axis) to the direction of the static magnetic field B ₀ . In general terms, the phase of the further high-frequency excitation pulse leads or lags the nuclear magnetic composite phase by 90° at the instant of the echo. In this way, another excitation pulse P4 rotates the vertical component into an equilibrium position or against the direction of the static magnetic field B ₀ . When more than one spin-echo signal is generated or when excitation pulses with different phases are used at the start of the measurement cycle, another high-frequency excitation pulse rotates the vertical component of the nuclear magnetism so that the static magnetic field B ₀ The phase required to make the direction or its opposite direction is easily derived. The final result is the pulse sequence used during the measurement cycle and the relaxation times T1 and T2 of the specimen.
Depends on. When a measurement cycle as shown in Fig. 4 is performed, the duration T1 of this measurement cycle is
It only takes 200ms. (selective) excitation pulse P1(+
x90° pulse), after tv1 (=25ms), +x
Apply 180° pulse P2. The subsequent echo signal F2 is sampled. Pulse P3 is -
x180° pulse, which follows 50ms after P2. Another high-frequency excitation pulse P4 must therefore follow 25 ms later. After this pulse P4 (+x, 90° pulse), a waiting time of 100 ms follows. Therefore,
The total duration of the measurement cycle is approximately 200ms. This is considerably shorter than the duration T of 500 ms to 1 second of the measurement cycle according to the state of the art (FIGS. 3a and 3b). After generation of a regular complete nuclear spin echo signal, the signal strength at the instant of the echo (in dynamic steady state) is calculated to be:

S ₁ = K・M ₀ {1−exp(−t4/T1)}exp(−2t1/T2)/1
-exp{-(t1+t2+t3)/T2}exp(-t4/T1) where _M0 is the induced nuclear magnetism at thermal equilibrium and K represents the instrument parameters. This signal strength S ₁ should be compared with the signal strength S ₂ obtained in the absence of another excitation pulse. After time t1
A 90° excitation pulse is followed by a 180° pulse. Therefore,
After this 180° pulse, an echo signal appears for a time t1. A time t2 elapses between the 180° pulse and the next excitation. Assuming that at this time t2 all components of the magnetization are out of phase due to the inhomogeneity of the magnetic field, the signal strength S at the instant of the echo
2 is given by the following equation.

S ₂ = KM ₀ {1-2exp(-t2/T1)+exp(-(t1+
t2)/T1}・exp(−2t ₁ /T ₂ ) Here, for the two extreme cases, the signal strength S ₁ and
Compare _S2 .

a If t4 (when another excitation is used) is larger than T1, then the following equation holds true.

S ₁ ≒ KM ₀ exp (−2t1/T2) A similar situation when no other excitation is used is t
Characterized by 2≫T1.

At this time, the following formula holds true for S ₂ .

A different situation occurs when S ₂ ≈KM ₀ exp (-2t1/T2) b t _i <<T1, T2 (i=1, 2, 3, 4). At this time, the following formula holds true for S ₁ .

S ₁ ≒KM ₀ t4/(T1/T2) (t1+t2+t3)+t4 On the other hand, S ₂ ≒KM ₀ (t2−t1/T1) A numerical example is given below for explanation. Select the value below. T1=T2=1000ms, t1=25ms, t2
=50ms, t3=25ms, t4=100ms. This is the case with additive excitation. In this case S ₁ ≒
0.5KM ₀ . Select the following values to find S ₂ .
t1=25ms, t2=175ms. Therefore, the same repetition time is obtained as when using additional excitation.
It is then found that S ₂ ≈0.15KM ₀ , which is considerably smaller than S ₁ . Obviously, when using a different excitation pulse, the image is not exactly the same as in the case of complete relaxation of the nuclear magnetism. However, in practice the waiting time between successive samples can be considerably shortened without significant signal loss, providing an image rich in information about the distribution of nuclear magnetism.

FIG. 5 shows a preferred variant of the method of the invention. The measurement cycle of this method consists of two cycles with substantially the same pulse sequence, but with different time intervals between the pulses. In this example, the high frequency 90° and 180° pulses in the first and second cycles correspond to the 90° and 180° pulses in FIG. time interval tv
1', tv2', t2', t3' are also tv1, tv2, t2,
Each corresponds to t3. However, time interval t4' is longer than time interval t4. A negative signal arises in the nuclear spin echo signal formed in the second cycle from the steady-state solution of the Broduch equation, resulting in a change in contrast (different intensity distribution) in the NMR image to be reconstructed of the object to be displayed. can lead to. Reduction of the measurement time mentioned above (periods T1 and T1
1 is the measurement time T in Figure 3a or Figure 3b.
should be chosen to be shorter than ) in addition to
Other information can be reproduced within the NMR image.
This information depends on the time intervals used in successive cycles.

It should also be noted that in the examples described above, 90° and 180° excitation pulses are used. Obviously the pulse at the beginning of the measurement cycle (P1, P1')
It is also possible to use other pulse angles (with or without selective excitation) and in another high-frequency pulse (P4, P4') at the end of the measurement cycle.

Preferably, pre-programmed computer means are used for selecting/adjusting a given pulse sequence and the associated time interval of the measurement cycle.
In one embodiment of the device 15 (FIG. 2), a central control device 45 has an input/output working terminal 5 for controlling data.
2 and pulse program generator 53 (see Figure 6)
It comprises a pre-programmed calculator 51 having the following. Output terminal 5 of pulse program generator 53
5 to the coil 3 via the bus 50 (see Figure 2).
Current generator 1 for a, 3b, 5, 7 and 11
9, 21, 23 and 25. Obviously, the output terminal 55 can also be connected directly to these current generators. Calculator (Philips type)
P857) can be programmed according to the program provided later as an appendix. Based on this program, the computer generates a pulse program generator (Nicolet Type 295B) using the program and control data input via the work terminal 52.
control. The set of instructions used in this program (third column of the program) is that of the pulse program generator 53 (excluding instruction JSA, which causes a jump to the start address).
Each term in the fourth column defines the time during which the output signal must be present at the output of the pulse program generator 53. The fourth column of the program (excluding the letter S) is 16
The state of the output side of the pulse program generator 53 is indicated by a decimal code. The fifth column states the address of the storage location. The symbol I in the sixth column indicates the presence of an interrupt. This interrupt can fetch additional functions as well as part of the code to be output to the output of the pulse program generator 53. This additional functionality includes, for example:
a) by loading the radio frequency signal generator 25 with a new waveform (when using 180° pulses instead of 90° pulses); b) by reversing the phase of the excitation pulse; or c) by indicating the start of a new pulse train. be. The program given in the appendix uses exclusively +y or -y pulses for 90° excitation pulses;
Exclusively +x or -x pulses are used for 180° excitation pulses.

Although we have discussed obtaining the nuclear spin resonance spectrum in Figure 3b under the condition that no steady gradient magnetic field exists during the measurement period, it is also possible to decide to take out the signal sample under the presence of a gradient magnetic field during the measurement period. . Dephasing the application of the gradient field after excitation by a dephasing delay τ _x (e.g. during the warm-up period) and then applying the two gradient fields to take the signal samples in the presence of a third gradient field. Applying a magnetic gradient field can also fill the 4-D matrix (K _x , K _y , K _z , t) with signal samples. However, the dephasing delay τ _x is different for each measurement cycle. In this way, after 4-D Fourier transform, x,
The desired spectrum is found as a function of y,z.

In the example above, each time the Fourier transform (Fourier
zeugmatography) to determine the nuclear magnetic distribution. However, it should be noted that the method of the present invention is
1983, pp. 73-89, a method based on the so-called projection reconstruction method is also covered.

Appendix PP0027: Name of pulse program 831006: Creation date of pulse program 100: Residence time expressed in microseconds (=4
*D11) The header of this pulse program contains a coded description of four high-frequency pulse sequences. first 3
These are those used in NMR imaging using driven equilibrium techniques. The last one is used for comparison purposes and does not use a driven balance technique. This last high frequency pulse sequence is commonly known as the saturation recovery high frequency pulse sequence. It is possible to select from one of the following high frequency pulse sequences:

(0) TZP high frequency pulse sequence: P(90)-TAU-P(180)-2*TAU-P(180)-
TAU-P (reset) - (D0 + 44) (1) VZP high frequency pulse sequence: P (90) - TAU - P (180) - 2 * TAU - P (180) -
TAU-P (reset)-P(180)-(D0+44) (2) VZP-VZP High frequency pulse sequence: P(90)-TAU-P(180)-2*TAU-P(180)-
TAU-P (reset)-P (180)-(D10+44)-P
(90)-TAU-P(180)-2*TAU-P(180)-
TAU-P (reset)-P(180)-(D0+44) (3) Saturation recovery high-frequency pulse sequence: P(90)-TAU-P(180)-2*TAU-P(180)-
TAU - (D0 + 44) 0: Start address of pulse program = 0 TZP (reset cycle), if sequence time = 144 + D0 microseconds = 1 VZP (set cycle), if sequence time = 144 + D0 microseconds = 2 VZP - VZP ( 2 * set cycle), sequence time = 288 + D0 + D10 microseconds = 3 saturation recovery, sequence time = 144 + D0 microseconds * * * * * * * * * * * * * Counter value: 256; C1 = time in echo End of number of samples; ＊ ＊ ＊ ＊ ＊ ＊ ＊ ＊ ＊ ＊ Duration value: (D0, D1, D2, ...D14) 256M; D0 = T (P (90)) - T ((Re) ) -44 microseconds 3M, 9M, 1M; Duration 1, 2 and 3 10M, 6M; Duration 4 and 5 7M, 1M, 1.2M, 8M; Duration 6, 7, 8 and 9 56M; D10=T(P(90))-T(P((re)set))-44 microseconds 25U, 19.53U; Duration 11, 12 D12=. 5*
D4/C1 3M, 2M; Duration 13, 14 End; * * * * * * * * * * * * Pulse menu 0 JUC D2 S8900E 7 ; Start for TZP, reset waveform generator 1 JUC D2 S8900E 50 ; For VZP Start, reset waveform generator 2 JUC D2 S8900E 94; Start for VZP-VZP, reset waveform generator 3 JUC D2 S8900E 136; Start for RHO-T2, reset waveform generator 4 JUC D2 S8900E 178; VZP/TZP Start cycle adjustment 7 NOP D1 S00E0E; Start/stop gradient magnetic field 8 NOP D13 S0000E; 9 NOP D1 S00E0E; Start/stop gradient magnetic field 10 NOP D1 S0000E; 11 NOP D7 S00E0E; Start/stop gradient magnetic field 12 NOP D3 S0000E; 13 NOP D4 S0013E; 90°+/-Y high frequency pulse 14 NOP D1 S00E0E; Start/stop gradient magnetic field 15 NOP D3 S0000E; 16 NOP D7 S00E0E; Start/stop gradient magnetic field 17 NOP D6 S5000E; Phase alternate Y 18 NOP D1 S00E0E ; Start/stop gradient magnetic field 19 NOP D3 S0000E ; 20 NOP D7 S00E0E ; Start/stop gradient magnetic field 21 LD1 D5 S00F8E C1 ; 180°+x high frequency pulse, load loop counter 22 NOP D1 S02E0A ; Gradient Start/stop magnetic field 23 NOP D8 S0200A; 24 NOP D1 S02E0A; Start/stop gradient magnetic field 25 NOP D3 S02000; 26 JUC D7 S02E00 29; Start/stop gradient magnetic field 27 NOP D11 S00000; NS＊T stay loop 28 IJ1 D11 S00000 31 ; 29 NOP D11 S00001 ; Sample pulse to ADC 30 JUC D11 S00001 27 ; 31 NOP D1 S00E0E ; Start/stop gradient magnetic field 32 NOP D8 S0000E ; 33 NOP D1 S00E0E ; Start/stop gradient magnetic field 34 NOP D3 S0000E ; 35 NOP D7 S00E0E ; Start/stop gradient magnetic field 36 NOP D5 S00F8E ; 180° + NOP D3 S0000E ; 41 NOP D14 S00E0E ; Start/stop gradient magnetic field 42 NOP D4 S00F3E ; 90°TZP (reset pulse) 43 NOP D1 S00E0E ; Start/stop gradient magnetic field 44 NOP D3 S0000E; 45 NOP D7 S00E0E; gradient magnetic field 46 NOP D5 S0000E; No high-frequency pulse 47 JSA D0 S0000E; Recycle 50 NOP D1 S00E0E; Start/stop gradient magnetic field 51 NOP D13 S0000E; 52 NOP D1 S00E0E; Start/stop gradient magnetic field 53 NOP D1 S0000E; 54 NOP D7 S00E0E ; Start/stop gradient magnetic field 55 NOP D3 S0000E ; 56 NOP D4 S0013E ; 90° +/- Y high frequency pulse 57 NOP D1 S00E0E ; Start/stop gradient magnetic field 58 NOP D3 S0000E ; 59 NOP D7 S00E0E; Start/stop gradient magnetic field 60 NOP D6 S5000E; Phase alternate Y 61 NOP D1 S00E0E; Start/stop gradient magnetic field 62 NOP D3 S0000E; 63 NOP D7 S00E0E; Start/stop gradient magnetic field 64 LD1 D5 S00F8E C1; 180°+X High frequency pulse, load loop counter 65 NOP D1 S02E0A; Start/stop gradient magnetic field 66 NOP D8 S0200A; 67 NOP D1 S02E0A; Start/stop gradient magnetic field 68 NOP D3 S02000; 69 JUC D7 S02E00 72; Start/stop gradient magnetic field Stop 70 NOP D11 S00000; NS*T stay loop 71 IJ1 D11 S00000 74; 72 NOP D11 S00001; Sample pulse to ADC 73 JUC D11 S00001 70; 74 NOP D1 S00E0E; Start/stop gradient magnetic field 75 NOP D8 S0000E ;76 NOP D1 S00E0E ; Start/stop gradient magnetic field 77 NOP D3 S0000E ; 78 NOP D7 S00E0E ; Start/stop gradient magnetic field 79 NOP D5 S00F8E ; 180°+X high frequency pulse 80 NOP D1 S00E0E ; Start/stop gradient magnetic field 81 NOP D9 S0000E ; 82 NOP D1 S00E0E ; Start/stop gradient magnetic field 83 NOP D3 S0000E ; 84 NOP D14 S00E0E ; Start/stop gradient magnetic field 85 NOP D4 S00F3E ; 90°TZP (reset pulse) 86 NOP D1 S00E0E ; Start/stop gradient magnetic field Stop 87 NOP D3 S0000E; 88 NOP D7 S00E0E; Start/stop gradient magnetic field 89 NOP D5 S0018E; 180° + 6 NOPs D1 S00E0E ; Start/stop gradient magnetic field 97 NOP D1 S0000E ; 98 NOP D7 S00E0E ; Start/stop gradient magnetic field 99 NOP D3 S0000E ; 100 NOP D4 S0013E ; 90°+/-Y high frequency pulse 101 NOP D1 S00E0E ; Gradient magnetic field 102 NOP D3 S0000E; 103 NOP D7 S00E0E; Start/stop gradient magnetic field 104 NOP D6 S5000E; Phase alternate Y 105 NOP D1 S00E0E; Start/stop gradient magnetic field 106 NOP D3 S0000E; 107 NOP D7 S00E 0E; slope Start/stop magnetic field 108 LD1 D5 S00F8E C1 ; 180°+X high-frequency pulse, load loop counter 109 NOP D1 S02E0A ; Start/stop gradient magnetic field 110 NOP D8 S0200A ; 111 NOP D1 S02E0A ; Start/stop gradient magnetic field 112 NOP D3 S02000 ; 113 JUC D7 S02E00 116 ; Start/stop gradient magnetic field 114 NOP D11 S00000 ; NS＊T stay loop 115 IJ1 D11 S00000 118 ; 116 NOP D11 S00001 ; Sample pulse to ADC 117 JUC D11 S00001 114; 118 NOP D1 S00E0E ; Start/stop gradient magnetic field 119 NOP D8 S0000E ; 120 NOP D1 S00E0E ; Start/stop gradient magnetic field 121 NOP D3 S0000E 122 NOP D7 S00E0E ; Start/stop gradient magnetic field 123 NOP D5 S00F0E ; 180°+X high frequency pulse 124 N OP D1 S00E0E ; Start/stop gradient magnetic field 125 NOP D9 S0000E ; 126 NOP D1 S00E0E ; Start/stop gradient magnetic field 127 NOP D3 S0000E ; 128 NOP D14 S00E0E ; Start/stop gradient magnetic field 129 NOP D4 S00F3E ; 90° TZP( 130 NOP D1 S00E0E ; Start/stop gradient magnetic field 131 NOP D3 S0000E ; 132 NOP D7 S00E0E ; Start/stop gradient magnetic field 133 NOP D5 S5018E ; 180°+X pulse, phase alternate Y 134 JUC D10 S0000E 1 ；VZP Go to cycle136 NOP D1 S00E0E; Start/Stop gradient field137 NOP D13 S0000E; 138 NOP D1 S00E0E; Start/Stop gradient field139 NOP D1 S0000E; 140 NOP D7 S00E0E; Start/Stop gradient field141 NOP D3 S0000 E ; 142 NOP D4 S0013E ; 90° +/-Y high frequency pulse 143 NOP D1 S00E0E ; Start/stop gradient magnetic field 144 NOP D3 S0000E ; 145 NOP D7 S00E0E ; Start/stop gradient magnetic field 146 NOP D6 S5000E ; Phase alternate Y 14 7 NOP D1 S00E0E ; Start/stop gradient magnetic field 148 NOP D3 S0000E ; 149 NOP D7 S00E0E ; Start/stop gradient magnetic field 150 LD1 D5 S00F8EC1 ; 180°+X high frequency pulse, load loop counter 151 NOP D1 S02E0A ; Start gradient magnetic field /stop 152 NOP D8 S0200A; 153 NOP D1 S02E0A; Start/stop gradient magnetic field 154 NOP D3 S02000; 155 JUC D7 S02E00 158; Start/stop gradient magnetic field 156 NOP D11 S00000; NS＊T stay loop 157 IJ1 D11 S 00000 160 158 NOP D11 S00001 ; Sample, pulse to ADC 159 JUC D11 S00001 156 ; 160 NOP D1 S00E0E ; Start/stop gradient magnetic field 161 NOP D8 S0000E ; 162 NOP D1 S00E0E ; Start/stop gradient magnetic field 163 NOP D3 S00 00E; 164 NOP D7 S00E0E ; Start/stop gradient magnetic field 165 NOP D5 S00F8E ; 180°+X high frequency pulse 166 NOP D1 S00E0E ; Start/stop gradient magnetic field 167 NOP D9 S0000E ; 168 NOP D1 S00E0E ; Start/stop gradient magnetic field 169 NOP D3 S0000E ; 170 NOP D14 S00E0E ; Start/stop gradient magnetic field 171 NOP D4 S00F0E ; No high-frequency pulse 172 NOP D1 S00E0E ; Start/stop gradient magnetic field 173 NOP D3 S0000E ; 174 NOP D7 S00E0E ; Start/stop gradient magnetic field 175 NOP D5 S0000E; no high-frequency pulse 176 JSA D0 S0000E; recycle 178 NOP D1 S00E0E; start/stop gradient magnetic field 179 NOP D13 S0000E; 180 NOP D1 S00E0E; start/stop gradient magnetic field 181 NOP D1 S0000E; 182 NOP D7 S0 0E0E; slope Start/stop magnetic field 183 NOP D3 S0000E ; 184 NOP D4 S0013E ; 90° +/-Y high frequency pulse 185 NOP D1 S00E0E ; Start/stop gradient magnetic field 186 NOP D3 S0000E ; 187 NOP D7 S00E0E ; Start/stop gradient magnetic field 188 NOP D6 S5000E ; Phase alternate Y 189 NOP D1 S00E0E ; Start/stop gradient magnetic field 190 NOP D3 S0000E ; 191 NOP D7 S00E0E ; Start/stop gradient magnetic field 192 LD1 D5 S00F8E C1 ; 180°+X high frequency pulse, loop counter Load 193 NOP D1 S02E0A; Start/Stop gradient magnetic field 194 NOP D8 S0200A; 195 NOP D1 S02E0A; Start/Stop gradient magnetic field 196 NOP D3 S02002; 197 JUC D7 S02E02 200; Start/Stop gradient magnetic field 198 NOP D11 S000 02; NS＊T Stay loop 199 IJ1 D11 S00002 202; 200 NOP D11 S00002; 201 JUC D11 S00002 198; 202 NOP D1 S00E0E; Start/stop gradient magnetic field 203 NOP D8 S0000E; 204 NOP D1 S00E0E; Start/stop gradient magnetic field 205 NOP D3 S0000E; 206 NOP D7 S00E0E; Start/stop gradient magnetic field 207 NOP D5 S00F8E; 180° + Start magnetic field /stop 211 NOP D3 S0000E; 212 LD1 D7 S00E0A C1; Start/stop gradient magnetic field 213 JUC D7 S00E08 215; Start/stop gradient magnetic field 214 IJ1 D11 S00E00 216; 215 JUC D12 S00E01 214; Sample, pulse ADC 216 NOP D1 S00E0E ; Start/stop gradient magnetic field 217 NOP D3 S0000E ; 218 NOP D7 S00E0E ; Start/stop gradient magnetic field 219 JSA D0 S0000E ; Recycle

[Brief explanation of the drawing]

Fig. 1 is a schematic diagram of the arrangement of the coil system of an apparatus for carrying out the method of the invention, Fig. 2 is a block diagram of the apparatus for carrying out the method of the invention, and Figs. 3a and 3b are simple examples of the current state of the art. 4 is a time diagram of an example of the method of the present invention, FIG. 5 is a time diagram of a preferred example of the method of the present invention, and FIG. 6 is a part of the apparatus for carrying out the method of the present invention. It is a block diagram. 1... Main coil (for static magnetic field generation), 2... Spherical surface on which the main coil is placed, 3... Coil for generating gradient magnetic field G _z , 5... Coil for generating gradient magnetic field G _y , 7... Gradient magnetic field Coil for generating G _x , 11... Coil for high frequency magnetic field, 15... Device for implementing the method of the present invention, 17, 19, 21,
23... Current generator for exciting the coil, 20
... human body, 25 ... high frequency signal generator, 27 ... high frequency signal detector, 28 ... demodulator, 29 ... sampling circuit, 31 ... processing means such as A/D converter, 3
3...Memory, 35...Arithmetic circuit, 37...Control device,
43...Display device, 45...Central control unit, 50...Bus, 51...Pre-programmed computer,
52...I/O work terminal, 53...Pulse program generator, 55...Output terminal.

Claims (1)

[Claims] 1. To measure the nuclear magnetic distribution within an area of an object, a uniform static magnetic field is generated, the area of the object is placed within this static magnetic field, and a) a high-frequency electromagnetic pulse is generated. to precess the magnetization of the nuclei in the object, thereby generating a resonant signal; b Then, after a preparatory period, in a measurement period that is divided into several sampling periods:
Periodically take several signal samples from this resonance signal (FID signal), c then each time after a waiting period, step a) above.
and b) several times (n') to obtain a group of signal samples, and in a nuclear magnetic distribution measurement method that obtains a distribution image of the nuclear magnetization induced from this, after extracting the sampling signal, a high frequency 180° pulse is generated. generating a nuclear spin echo signal to counteract the effect (phase shift) imposed on the nuclear magnetization by the gradient magnetic field, and then generating another radio frequency excitation pulse such that a nuclear spin echo signal is present;
and after said another RF excitation pulse, after a period of time T, a similar cycle of the RF excitation pulse and the associated gradient magnetic field is completed, reducing the at least one pulse interval during the first cycle to a time period T during the second cycle. A nuclear magnetic distribution measurement method characterized by making the pulse interval different from the corresponding pulse interval. 2. The nuclear magnetic distribution measuring method according to claim 1, characterized in that said another high-frequency excitation pulse is generated when the maximum value of the nuclear spin echo signal occurs. 3. The nuclear magnetic distribution measuring method according to claim 2, characterized in that the phase of another high-frequency excitation pulse is advanced by 90 degrees from the phase of nuclear magnetization at the instant of echo. 4. The method for measuring nuclear magnetic distribution according to claim 2, characterized in that the phase of another high-frequency excitation pulse is delayed by 90 degrees from the phase of nuclear magnetization at the instant of echo. 5. The method for measuring nuclear magnetic distribution according to claim 3 or 4, characterized in that the other high-frequency excitation pulse is a 90° selective excitation pulse. 6. During the preparation period, the first or first and second gradient magnetic fields are applied, the directions of the gradients are perpendicular to each other, and the direction of the gradient during the measurement period is perpendicular to the direction of the gradient of the gradient magnetic field that exists during this preparation period. applying another gradient magnetic field extending in a direction such that the integral of the strength of at least one gradient field during the preparation period has a different value each time steps a) and b) are repeated, and the nuclear spin density 6. The nuclear magnetic distribution measuring method according to claim 1, wherein the distribution is extracted from a group of signal samples after Fourier transformation. 7 Applying at least one gradient magnetic field in the preparatory period, the integral of the intensity of which in the preparatory period has a different value each time when steps a) and b) are repeated, and samples the position-dependent nuclear spin resonance spectrum as a signal. Claims 1 to 5 are characterized in that the method is extracted from the group after Fourier transformation.
The method for measuring nuclear magnetic distribution according to any one of paragraphs. 8 After generating the first high-frequency electromagnetic pulse and before applying the gradient magnetic field, including a dephasing period, which dephasing period has a different value each time steps a) and b) are repeated. The method for measuring nuclear magnetic distribution according to claim 7, characterized in that: 9 In order to generate the first high-frequency electromagnetic pulse, one selective gradient magnetic field (selective excitation) is applied during the preparation period, two or three gradient magnetic fields are applied,
The nucleus according to claim 7, wherein the directions of the gradients are perpendicular to each other, and a nuclear resonance spectrum that depends on the position within the cross section or volume of the object is obtained by three-dimensional or four-dimensional Fourier transformation. Magnetic distribution measurement method. 10 In order to determine the nuclear magnetic distribution within the region of an object, a means for generating a uniform static magnetic field, b a means for generating high-frequency electromagnetic radiation, c a means for generating a gradient magnetic field, d a) and sampling means for sampling the resonance signal generated by the means specified by b); e processing means for processing the signal provided by said sampling means to form a nuclear magnetic distribution; and f at least the terms b) or e. ) to generate some resonant signals,
a nuclear magnetic distribution measurement apparatus comprising control means for conditioning, sampling and processing, the control means comprising preprogrammed computer means for generating and supplying control signals to the means for generating high frequency electromagnetic radiation; , the above control signal is
Adjustable pulse trains of 90° and 180° excitation pulses can be generated, with the period elapsed between two recently generated 180° excitation pulses being the same as the most recent 180° pulse and the next 90° excitation. 1. A nuclear magnetic distribution measuring device characterized in that the period is twice as long as the period that elapses between pulses.