Note: Descriptions are shown in the official language in which they were submitted.
Z4~
OO 44.5~ 365
IMPROVEMENTS IN AND RELATING TO
MAGNETIC RESONANCE IMAGING
The present invention relates to improvements
in and relating to magnetic resonance imaging (MRI)
apparatus and methods, and in particular to apparatus
and methods for diagnostic imaging and to contrast
agents for use in such methods.
MRI is a diagnostic technique that has become
particularly attractive to physicians as it is
non-invasive and does not involve exposing the
patient under study to potentially harmful radiation,
such as for example the X-radiation of conventional
radiography.
This technique however suffers from several
serious draw-backs, including in particular the
expense of manufacture and operation of the MRI
apparatus, the relatively long scanning time required
to produce an image of acceptable spatial resolution,
and the problem of achieving contrast in the magnetic
resonance (MR) images between tissue types having
the same or closely similar imaging parameters,
for example in order to cause a tissue abnormality
to show up clearly in the images.
The expense of manufacture and operation
of an MRI apparatus is closely associated with
the field strength that the primary magnet in the
apparatus is required to generate in order to produce
images of acceptable spatial resolution within
an acceptable time (the~image acquisition time~.
In general, magnets capable of generating field
strengths of 0.1 to 2 T have been used and image
acquisition times are commonly of the order of
10-30 minutes. For relatively low field strengths
~3(~
-- 2
of up to 0.15 T, resistive magnets (generally adiacent
coaxial metal coils) may be used but the energy
requirement (and as a result the heat generation)
of such resistive magnets is very high. Thus a
0.1 T magnet will reauire about 30kW electric power.
For higher fields, superconducting magnets are
conventionally used. The selection of the appropriate
magnetic field strength involves balancing various
factors; thus higher field results in a better
signal/noise (S/N~ ratio and hence better spatial
resolution at a given S/N value, but also in greater
manufacturing and operating expense and in poorer
tissue contrast. There is therefore a demand for
MRI apparatus and techniques capable of achieving
improvements in S/N ratio, especially if such apparatus
would enable lower field magnets to be used without
undue loss in spatial resolution.
The long image acquisition times generally
result from the need to perform a large number
(e.g. 64-1024) of pulse and detection sequences
in order to generate a single image and in the
need to allow the sample under study to reequilibrate
between each sequence.
The degeneracy of the spin states of nuclei
with non~zero spin, e.g. lH 13C 19F t
lost when such nuclei are placed within a magnetic
field and transitions between the ground and excited
spin states can be excited by the application of
radiation of the frequency (~O) corresponding to
energy difference E of the transition (i.e.
~o=E). This frequency is termed the Larmor frequency
and is proportional to the strength of the applied
field. As there is an energy difference between
the spin states, when the spin system is at equilibrium
the population distribution between ground and
excited spin states is a Boltzmann distribution
~3~ 8
and there is a relative overpopulation of the ground
state resulting in the sp;n system as a whole posses-
sing a net magnetic moment in the field direction.
This is referred to as a longitudinal magnetization.
At equilibrium the components of the magnetic moments
of the individual non-zero spin nuclei in the plane
perpendicular to the field direction are randomized
and the spin system as a whole has no net maqnetic
moment in this plane, i.e. it has no tranverse
magnetiæation.
If the spin system is then exposed to a relatively
low intensity oscillating magnetic field perpendicular
to the main field produced by radiation at the
Larmor frequency, generally radiofrequency (RF)
radiation, transitions between ground and excited
spin states occur. If the exposure is for a relatively
short duration then the resultant magnitudes of
the longitudinal and transverse magnetizations
of the spin system are functions of the exposure
duration which oscillate about zero at the Larmor
frequency and are 90 out of phase with each other.
Thus, from equilibrium, a pulse of duration (2n+1)1
/2~o ~a so-called 90 pulse when n is even and
a 270 pulse when n is odd) leaves the system with
maximum transverse magnetization (of magnitude
proportional to the initial longitudinal magnetization
at equilibrium~ and no longitudinal magnetization,
a pulse of duration (2n+1) ~ /~o (a 180 pulse)
leaves the system with inverted longitudinal magnetiza-
tion and inverted transverse magnetization (and
hence from equilibrium no transverse magnetization),
etc.
When the pulse is terminated, the oscillating
magnetic field produced by any resulting net transverse
magnetization can induce an oscillating electrical
signal (of angular frequency ~O) in a detector
~3~
- 4 - 20208-1346
coil having its axis arranged perpendicular to the main field
dixection. For this purpose the transmitter used to emit the
pulse can also be used as a detector.
Induced magnetic resonance signals, hereinafter termed
free induction decay (FID) signals, have an amplitude proportional
to the transverse magnetization (and hence generally to the
original population difference between ground and excited spin
states).
If the nuclei of the spin system experienced an
entirely uniEorm magnetic field, the FID signal would decay due
to spin-spin interactions at a rate with a characteristic time of
T2, the transverse or spin-spin relaxation time. Elowever, due to
local field inhomogeneities, the nuclei within the spin system
will have a spread of Larmor frequencies and decay of transverse
magnetization is more rapid, having a characteristic time of T2*
where l/T2* = 1/T2 + l/Tinh, Tinh representing the contribution
due to field inhomogeneities. T2 itself can be determined using
spin-echo imaging in which, after the decay of the FID signal
(usually following a 90 pulse) the system is exposed to a 180
pulse and an "echo" signal is generated, the decay in the amplltude
of the echo being governed primarily by T2 as, with the inversion
of the transverse magnetization for the individual nuclei, the
field inhomogeneities referred to above cause transverse magnetiz-
ation to build up to a maximum at time TE/2 after the 180 pulse
where the time between the previous maximum transverse magnetiz-
ation and the 180 pulse is also TE/2.
13~t~Z4~
- 4a - 20208-1346
To generate different images, different pulse and FID
detection sequences are used. Perhaps the simplest is saturation
recovery (SR) where the FID
13~Z~8
siqnal is determined after a single 90 initiating
pulse. The signal strength is dependent upon the
maqnitude of the longitudinal magnetization before
the pulse, and hence on the nuclear density and the
extent to which the system reequilibrates in the
time (TR) between succe~sive initiating pulses. In
spin-echo imaging, for example multiple-echo imaging,
the pulse and detection sequence may be: initiating
90~ pulse (at time 0), FID detection (following the
initiating pulse~, 180 pulse (at time TE/2), detection
of 1st echo (at time TE), 180 pulse (at time 3TE/2),
detection of 2nd echo (at time 2TE)..., initiating
pulse for the next sequence (at time TR), etc. In
this technique, a TR is selected which is sufficient
for a reasonable reequilibration to occur in the
period between successive initiating pulses.
As is explained further below in connection
with the example of two dimensional Fourier transforma-
tion (2DFT) image generation, in order to generate
a single image with adequate spatial resolution,
it is necessary to perform a large number (e.g. 64-
1024) of separate pulse and detection sequences.
Since TR has in principle to be large with respect
to Tl, the characteristic time for relaxation of
the excited system towards the equilibrium Boltzmann
distribution between ground and excited spin states,
to permit longitudinal magnetization to build up
between successive pulse sequences so as to avoid
the FID signal strength decaying in successive pulse
sequences, the total image acquistion time is generally
relatively large. Thus, for example, TR may convention-
ally be of the order of seconds and the image acquisition
time may be of the order of 10 30 minutes.
Certain so-called fast imaging (FI) techniques may
be used to accelerate reequilibration and so reduce
image acquisition time; however they inherently result
13~LZ~8
-- 6
in a reduction in the ~/N ratio and/or contrast hence
in poorer image quality. The FI techni~ue invo]ves
for example exciting the spin system with a less
than 90 pulse and thus the difference between ground
and excited spin state populations is only reduced
rather than eliminated (as with a 90 pulse) or inverted
and so reattainment of equilibrium is more rapid.
Nevertheless, the transverse magnetization generated
by the less than 90 pulse is less than that for
a 90 pulse and so FID signal strength and thus S/N
ratio and the spatial ~esolution in the final image
are reduced.
The long image acquisition time in conventional
MRI thus significantly detracts from the attractiveness
of MRI for mass or routine diagnostics screening
and for all forms of diagnostic imaging where it
is necessary to build up a three-dimensional image
by imaging successive adjacent sections through the
patient.
There is thus a demand for MRI apparatus and
techniques which allow reduction in image acquisition
time without undue loss in resolution or contrast.
The third problem mentioned above, that of
achieving adequate image contrast between different
tissue types, has been addressed in a variety of
ways. Using different pulse and detection sequences
and by manipulation of the acquired data, MRI can
be used to generate a variety of different images,
for example saturation recovery (SR), inversion recovery
(IR), spin echo (SE), nuclear (usually proton) density,
longitudinal relaxation time (T1) and transverse
relaxation time (T2) images. Tissues or tissue abnorma-
lities that have poor contrast in one such image
often have improved contrast in another. Alternatively,
imaging parameters (nuclear density, Tl and T2) for
-- 7 --
tissues of interest may be altered by administration
of a contrast agent. Thus many proposals have been
made for the administration of magnetically responsive
materials to patients under study (see for example
EP-A-71564 (Schering~, US~A-4615879 (Runge), ~IO-A-
85/02772 (Schroder) and WO-A-85/0~330 ~Jacobsen)).
Where such materials, generally referred to as MRI
contrast agents, are paramagnetic (for example gadolinium
oxalate as suggested by Runge) they produce a significant
reduction in the Tl of the water protons in the zones
into which they are administered or at which they
congregate, and where the materials are ferromagnetic
or superparamagnetic (e.g. as suggested by Schr~der
and Jacobsen) they produce a significant reduction
in the T2 of the water protons, in either case resulting
in enhanced (positive or negative) contrast in the
magnetic resonance (MR) images of such zones.
The contrast enhancement achievable by such
agents is limited by a number of factors. Thus such
contrast agents cannot move the MRI signal intensity
(Is) for any tissue beyond the maximum (Il) and minimum
(Io) intensities achievable for that tissue using
the same imaging technique (e.g. IR, SR, SE, etc.~
in the absence of the contrast agent: thus if "contrast
effect" is defined as (Is-Io)/~ Io), contrast agents
can serve to alter the "contrast effect" of a tissue
within the range of 0 - 1. However to achieve contrast
improvement an adequate quantity of the contrast
agent must be administered to the subJect, either
directly to the body site of interest or in such
a way that the natural operation of the body will
bring the contrast agent to that body site.
- ~ -
There is therefore a continuing demand for
techniques capable of achieving enhanced contrast,
especially in tissues difficult to target with conven-
tional contrast agents and a technique capable of
achieving contrast effects greater than 1 would be
particularly desirable.
We now propose to utiliæe the spin transition
coupling phenomenon known in conventional nmr spec-
troscopy as the Overhauser effect to amplify the
Bolt~mann population difference due to relaxation
of the nuclear spin system producing the MR image
by exciting a coupled esr transition in a paramagnetic
species naturally occurring in or introduced into
the sample being imaged, generally but not essentially
a human or animal sub~ect.
The MRI apparatus for use according to this
technique requires a second radiation source for
generating the radiation capable of stimulating such
an esr transition as well as the first radiation
source for generating the radiation used to stimulate
the nuclear spin transition. In general, at the
magnetic fields that would normally be used with
such apparatus, the first and second radiation sources,
will be radiofrequency (RF) and microwave (MW) sources
respectively.
Thus in one aspect, the present invention provides
a magnetic resonance image generating apparatus compris-
ing a first radiation source capable of emitting
a first radiation of a frequency selected to excite
nuclear spin transitions in selected nuclei in a
sample being imaged and means for detecting free
induction decay signals from said selected nuclei,
characterised in that said apparatus further comprises
a second radiation source capable of emitting a second
~3~`~Z4~3
. 20208-13~6
_ g --
radl~tion of a fre~uency selected to exc~te electron
spin transitions ~oupled to the nuclear ~pin tranfiltion~
of at least some of said ~elected nuclei.
In a further aspect, the invention also provides
a metho~ of genera~ing a magnetic resonan~e
image of a sample, ~ald method comprising introducinq
into ~aid sample a con~ra~t medium comprls~ng a paramaqnetlc
mater~l; exposing said ~ample to a fi~st radiation
of a fre~uency ~electea to excite nuclear spln tran~ltlons
in sele~ted nu~lei in ~ala ~ple and aetecting ~ree
inauction decay sign~ls ~rom ~aid sample, characterised
in that ~aid paramagnetic material ha& an electron
~pln resonance transition baving a linewidth of
10 4 Tesla or le~ and ln tha~ ~aid method ~urther
comprl~es expo~inq ~aid 8ample t~ a ~econa radlation
o~ a ~re~uency wh~ch ex~ites e~ectron spin tran6ition~
c~pled to nu~lear spin tr~n~itions o~ at least some
of said nuclei, said se~ona radiation being of ~
frequency selecte~ to excite a said electron spln
resonanc~ tran~ition o~ sa$d paramagnetic materi~l.
In the method of the ln~ention the ~ampl~ is
conveniently ~xpo~ed to ~ seriea o~ pul~e ~equen~e~
of ~aid fir~t radiation and exposed to the ~e~ond
radiatlon for at leAst part of eaah pulse sequence,
.e. durlng at le~t part af the period ~etween ~he
inltlal pulse~ of ad~acent ~aid sRquences. Preferably
expo~ure to the ~econd raaiation will be for some,
the ma~or part or all of th~ p~rloa during which
no magnetic field grDdient ls lmpo~ed on the sample.
Conveniently there~ore the ~econd radiation may be
~pplied following FID ~ig~al d~termination in eacb
pul~e seq~ence.
.. ~,~ .
~3~Z~3
20208-1346
- 9a -
It w~ll be appreciated that for certain imaging
~echnique~, parti~ularly s~turation recover~ (S~)
~ach ~pul~e sequence" ma~ only involve o~e pul~e
of the first radiatîon while in other MR imaqing
techni~ues each pulse ~equ~nce may involve several
pul~es of the fir~t radiaSi~n.
The magnetic resonance (MR) im~qe o~ tbe sam~le
can be generated n the conventional manner from
~3~ 8
-- 10 -
the detected ~ID siqnals. In particular it is not
necessary to detect comparat;ve signals from the
sample while this is not exposed to the second radiat;on.
Thus generally the apparatus of the invention will
comprise means, generally a computer, for transforminq
the detected FID signals into MR images, these means
being arranged to generate such images using only
signals detected following emission of both first
and second radiations by the radiation sources.
In conventional nmr spectroscopy, it has long
been known that if a sample comprising a paramagnetic
species and a species containing non-zero spin nuclei,
for example sodium dissolved in ammonia, is placed
in a strong magnetic field and an esr transition
of the paramagnetic species (sodium) is saturated,
then peaks in the nmr spectrum of the other species
can be very strongly enhanced due to coupling between
the electron and nuclear spin transitions. The effect
has been termed the Overhauser effect, or dynamic
nuclear polarization as exciting the esr transition
drives a nuclear spin system at equilibrium towards
a new equilibrium distribution with a relatively
higher excited state population. In the present inven-
tion, this effect is operated not as in conventional
spectroscopy to generate a strong peak in an nmr
spectrum but instead to amplify population difference
due to relaxation of an excited nuclear spin system.
The amplified population difference achieved
using the method and apparatus of the invention may
be utilized beneficially in a number of different
ways.
Thus to achieve the same S/N ratio and as a
result the same spatial resolution, a lower strength
main magnetic field (e.g. a lower power and thus
more economic primary magnet) and/or a shorter sequence
~3~ 8
repetition period T~ (and hence a shorter image acquisi-
tion time~ may be used. Alterna~ively, with no reduction
or lesser reductions in main field strength an increase
in signal strength (corresponding to a contrast effect
of greater than 1~ can be obtained, the maximum increase
in signal strength being obtained when TR is selected
to permit the spin system to reach the new equilibrium
between the pulse seauences.
Where only a portion of the nuclei whose spin
transitions produce the FID signal (hereinafter the
"resonating nuclei") couple with the unpaired electrons
of the paramagnetic species, for example due to low
concentration or non-uniform distribution of the
paramagnetic species in the volume being imaged,
the operation of the method and apparatus of the
invention will also result in contrast enhancement
in the image - thus the FID signal from the resonating
nuclei coupling with the unpaired electrons will
be enhancea relative to the signals from the non-
coupling nuclei. Where the paramagnetic species
is either naturally abundant in specific tissues
only or is administered in a contrast agent so as
to congregate in such tissues, the operation of the
invention will therefore allow generation of images
in which the contrast enhancement of these tissues
is high. It should be noted however that where the
power leve:L of the second radiation or the concentration
of the paramagnetic material is particularly low
it is possible for MR image intensity to be reduced
rather than enhanced. Even in such cases however
the modified contrast achieved in the resulting MR
images may be of interest.
-
As mentioned above, the pàramagnetic substancepossessing the esr transition which couples with
the nmr transition of the resonating nuclei may preferably
either be naturally present within the sample or
~3~Z~3
, ~
may be administered thereto in a contrast agent.
Coupling with the resonating nuclei may be either
scaler coupling with resonating nuclei within the
same molecules as the unpaired electrons or dipolar
coupling with resonating nuclei, ~enerally water
protons in the body fluids, in molecules in the environ-
ment of the paramagnetic centres.
Electron spin systems do occur naturally in
the body, e.g. in substances synthesized in certain
metabolic pathways such as the oxidation chain in
the cell mitochondria.
Insofar as administered contrast agents are
concerned however, in one embodiment of the invention
there may be used a contrast medium which contains
both the resonating nuclei and the substance posseæsing
the desired electron spin transition, and in a further
embodiment the substance possessing the desired electron
spin transition may itself also contain one or more
of the resonating nuclei. This is especially preferred
where the resonating nuclei are rarely abundant in
the sample being imaged, for example where the resonating
nuclei are C or 19F nuclei where scalar coupling
will be important in the amplified FID. Using such
a contrast agent, the FID signal will derive predominantly
from body ~;ites containing the contrast agent thereby
facilitating imaging of specific tissues or organs.
Alternatively, and generally more preferably,
the contrast agent may contain a paramagnetic centre
which undergoes dipolar coupling with resonating
nuclei naturally occurring in the sample, e.g. in
body tissue, or more specifically with resonating
protons in water molecules in the sample.
In the method of the invention, selection of
the esr system which couples with the resonating
~3(31~-
nuclei is particularly împortant where the imaging
is to be performed on a live subiect. In particular,
for efficient amplification of the nuclear spin popula-
tion difference, the second radiation should be such
as to maintain the electron spin system in an excited
state, preferably the saturated state. ~owever, in
in vivo imaging it is desirable to minimize the exposure
of the patient to the second radiation (generally
MW) in order to avoid unwanted heating effects and
it is therefore desirable to select an esr transition
with long transverse and longitudinal relaxation
times, T2e and Tle, to allow saturation of the transition
to be achieved without undue heating of the sample.
Since the line widths of esr transitions in
the esr spectrum are proportional to T2e 1, the bandwidth
required for the second radiation that is used to
saturate the esr transition will be smaller where
the transition corresponds to a narrow line in the
esr spectrum and a long transverse relaxation time
i8 therefore desirable. Similarly since the second
radiation absorption required to maintain saturation
is higher for a shorter Tle, a long longitudinal
relaxation time is desirable.
Particularly preferably, the substance possessing
the esr transition excited by the second radiation
will be a paramagnetic material whose esr spectrum
consists of a single narrow line or a set of closely
adjacent narrow lines (for example resulting from
hyperfine splitting of a single transition under
the effect of neighbouring non-zero spin nuclei within
the structure of the paramagnetic substance). Where
the esr spectrum contains a reasonably small number
of lines it will, as discussed below, be possible
simultaneously to excite many or all of the corresponding
transitions.
13Vl~
- ]4 -
~ onventional pa~amagnetic M~I contrast agents,
such as the gadolinium compounds (e.g. Gd-DTPA) suggested
by Scherinq (EP~A-71564), have large spectral linewidths
and would not generally be selected since they are
highly likely to require unacceptable microwave heating
of the sample in order to achieve any significant
amplification of the FID signal. Generally therefore
where a contrast medium is to be used as the source
of the esr transition, it should preferably have
a stimulable esr transition having a line width (i.e.
full width at half maximum in the absorbtion spectrum!
of the order of 1 Gauss or less, preferably 100 milli~auss
or less, and especially preferably 50 milliGauss
or less. If the esr spectrum contains a plurality
of lines it is furthermore preferred that the total
number of these lines be small, for example 2-lO,
preferably 2 or 3, and/or that the lines or a majority
thereof should be separated by not more than about
30 MHz at the operating field of the MRI apparatus
in order that several or all of the corresponding
esr transitions may be excited.
In general, to avoid hyperfine splitting, the
paramagnetic material will most desirably be a molecule
containing no non-zero spin nuclei, or containing
non-zero spin nuclei only at positions remote from
the paramagnetic centre. Conveniently, the molecule
may have the atoms near to the paramagnetic centre
predominantly selected rom zero nuclear spin isotopes
or from elements for which the natural abundance
of non-zero spin nuclear isotopes is low. Such selection
may include elements in which the natural abundance
of spin = ~ nuclei is low and isotopes such as 12C,
3 S, l4Si and 160 may for example be used to build
up the molecular structure adjacent to the location
of the unpaired electron. Alternatively, paramagnetic
materials having nuclei which give rise to hyperfine
splitting of the esr transition but with very small
coupling constants may be considered.
13U ~Zf~13
~ s an example, the use of a paramagnetic material
such as the stable free radical anion chloranil semi-
quinone -anion radical might be contemplated. ~here
a stable free radical is to be used however it may
be necessary to generate the stable free radical
species from a precursor compound before administration
of the contrast medium, e.g. by exposure of the contrast
medium to radiation or heat or by chemical treatment.
One particularly interesting group of stable
free radicals are the nitroxide stable free radicals
of which many have been suggested in the literature
for use as spin labels or as paramagnetic contrast
agents for conventional MRI. Moreover, several
of these compounds are readily available commercially,
for example from Aldrich. The nitroxide stable
free radicals are of particular interest as their
toxicities and pharmacokinetics have been studied
and show the compounds to be suitable for in vivo
M~ and as the esr line widths, especially for
compounds in which the atoms adjacent to the NO
moiety are fully substituted (i.e. carry no protons),
are adequately small at the concentrations re~uired
to give contrast enhancement.
As the nitroxide stable free radical, there
may conveniently be used a cyclic nitroxide wherein
the NO moiety occurs in a 5 to 7-membered saturated
or ethylenically unsaturated ring with the ring
positions ad~acent to it being occupied by doubly
saturated carbon atoms and with one of the remaining
ring positions being occupied by a carbon, oxygen
or sulphur atom and the remaining ring positions
being occupied by carbon-atoms.
Preferred nitroxides may be represented by
the formula (I)
~l3~
- 16 -
wherein Rl to R4 may represent lower (for example
Cl 4) alkyl or hydroxyalkyl groups and Rl may also
represent carboxy substituted Cl ~0 alkyl groups
and R2 may also represent a higher (e.g. ~5 20)
alkyl group or a carboxy substituted Cl 20 alkyl
group, and X represents an optionally substituted,
saturated or ethylenically unsaturated bridging group
having 2 to A atoms in the backbone of the bridge
one of the backbone atoms being carbon, oxygen or
sulphur and the remaining backbone atoms being carbon.
In formula I, the moieties CRlR2 and CR3R4
are preferably the same. Particularly preferably,
Rl to R4 are all methyl groups.
In formula I the optional substitution on X,
which preferably is an optionally mono-unsaturated
C2 3 chain, may for example take the form of halogen
atoms or oxo/ amino, carboxyl, hydroxy or alkyl groups
or combinations or derivatives thereof such as for
example amide, ester, ether or N-attached heterocyclic,
e.g. 2,5-dioxo-pyrrolidino, groups. Many examples
of substituted X groups are described in the literature
mentioned below.
The nitroxide molecule may if desired be bound
to a further substance, such as for example a sugar,
polysaccharide, protein or lipid or to other biomole-
cules, for example to enhance the blood pooling effect
- 17 -
or the t;ssue- or organ- targetting ability of the
nitroxide stable free radical.
Thus for example CA-A-1230114 (Schering) describes
nitroxide stable free radicals (for use as MRI contrast
agents) of formula II
y-co-~
~ (C~
S ~ N ~
R~ I ~t
wherein B is a protein, sugar or lipid residue
or a group -NRgRlo, --- is a double or single
bond, Y is -(CH2)n~ or if __- is a single bond
also -NH CO(CH2)n-, n is a number from 0 to 4,
m is a number from 0 to 2, Rg and Rlo are hydrogen
or alkyl optionally substituted by hydroxy, acyloxy
or alkylidenedioxy (both Rg and Rlo however cannot
simultaneously be hydrogen or unsubstituted alkyl),
R5 and R7 are alkyl and R6 and R8 are optionally
hydroxy substituted alkyl.
Furthermore, W087/05222 (MRI Inc.~ describes
nitroxide stable free radicals (again for use as
MRI contrast agents) of formula III
A 7~/~'5
R ,~
R,~
- 18 -
(wh~n Rll to R14 are each optionally hydroxyl substi-
tuted Cl 4 alkyl, A is ~2 4 alkylene or alkenylene,
-CH -~-CH2- or -C~2-S-CH2- and R15
E-COO~ where E is Cl ~ alkylene and M is NH4, Na
or K or R15 is -N(Alk)3 , Hal where ~al is a halogen
atom and Alk is Cl_8 alkyl optionally substituted
by hydroxy or esterified hydroxy), of formula IV
R,~ coo ~
R~7~ ~ (IV)
. R,8 o. R2 o
(wherein M is as defined above, R16, R17 and R18
are alkyl, cycloalkyl, heterocyclic aliphatic, carbo-
cyclic aryl or heterocyclic aryl and R19 and R20
are carbocyclic or heterocyclic aryl) and of formula V
A ~
~ a (V)
~1~ O, R~
( in Rll, R12, R20 and A are as defined above
21 R2~ are -(Cl_8 alkylene)-R2 where R
is hydrogen, R15, NH2, NHR15 or NRll 12 15
is as defined above).
Still further nitroxide stable free radicals for
use as MRI contrast agents are disclosed in WO87/01594
(Amersham International PLC) and in the references
cited therein. The nitroxides disclosed by Amersham
~3~ 8
-- ].9
are bound, optionally through the agency of linker
molecules, to polysaccharides such as dextran, starch
and cellulose.
A nitroxide stable free radical of formula VI
o
r
~ ~ (VI)
is disclosed by Alcock et al in Tetrahedron 33 (1977)
2969 - 2980.
Nitroxide stable free radicals of formula VII
(VII)
O'
~wherein Z is a hydroxyl, ethoxy or a substituted
amino group) are disclosed by Golding et al. in Synthesis
7 (1975) 462-463.
Nitroxide stable free radicals of formulae
VIII and IX ~ Cc~3~ 3
(VIII) ~ ~IX)
O~ O
, .
(wherein R24 is COOH or CONHCH(CH2OH)CHOHCH20H) and
their pharmacokinetics are discussed by Eriksson
et al. in J. Pharm. Sci. _ (1988) 97-103.
~3~
20208-134S
-- ~U - .
Moreover. nitroxide ~table free radical~ ar~
discussed generally b~ C.F. Chignell in "The Application
of Electron ~pin Resonance and Spin-labelling in
B~ochemistry and ~armacology", page~ 1 to 6, a publica-
tion w~ich indicate~ at page 6 that the following
nitroxide stable free radical~ are commercially available
~ro~ Aldrl~h:
_~o~H~ ~o~t~q
o~
t'J ~N ~ ~ N
I
O O.
other pAramaqnetic m~teriAls, the u~e o~ whi~h
may be contemplated include the 3,5-diçhloro-~,4,6-
tri ~hydroxyalkoxy or tri~hydroxyalkyl)s~lyl)-phenoxy
radical~ and the di(t~l~hydroxyaLkyl)~ilanyl)cyclo-
butadienoqulnoneS, in either ~a~e the hydroxyalkyl
moietie~ conveniently ~ontaining from 2 to 4 carbon
atoms and being for example 2-hydroxyethyl, 2,~-dlhy-
droxypropyl or 3,4-dihydroxybutyl ~roups.
~ n a still further ~pect the inventlon alsa
provide~ the use of a phy~iologically tolerable para~ag-
netic material, e.g. A s~able f~ee radical, havlnq
an electron ~pin re~onance transition wlth a linewidth
of l Gaus~ (lO-~ ~sla) or les~ for the manufact~r2
o~ a contrast meaium for use in a method of imag~ng
o~ a human or animal ~o~y h~cording to the pres~nt
invention.
3"' '
~3~12~8
20208-1346
It will be appreclated that where references
are made here n to the limit~ ~or e~r linewidths
thes~ will be the linewidth~ at imagîng condition~,
e.g. at the lmaged sites. Partiaularly preferab~y
however the linewid~h criterla will be ~atis~ied
at the local concentrat~on limits mentloned below.
The cQntrast m~dium may contain, be~lde~ t~e
p~r~magnet~c material, formulation aid~ suoh as are
aonventlonal ~or therapeutic and diagnostic compo81t~0ns
in human or vete~inary medicine. ~hus the medi~
may for example inalude 801u~11i2ing agents, emulsifiers,
vi8co~ity enhancer~, buffers, etc. ~he media may
be in ~orm~ ~ultable for parenteral (e.g~ lntravenous)
or enteral ~e.~. oral) application, for example for
appllcation directl~ lnto bDdy cav$t$e~ having external
e wape duct& ~u~h a~ th~ dlge~tive tract, the bladder
and ~he uterus), o~ ~or injection or infuslon into
the card~ov~cular sy~tem. However, 801utions, suspen-
~ion~ ~nd dispers~on~ in phy~iolo~ically toler~ble
media will generally b~ pre~er~ed.
For u~e in in ~ivo dihgno~tic ~maging, the
~ontrast medium, which preferably will be ~ubstAntially
i~o~onlc, may conveniently be ~dmlniatered At a conoen-
~ration ~uf~eicient to yleld a l~M ~o lOmM aoncentration
of the paramagnetl~ sub~tance at the image zone
however the preciqe con~entration and dosage will
of cour~e depend upon a range of ~ctor8 ~uch as
x
~,' .
Z4E~
20208-1346
- 22 -
toxicity, the or~an targe~tin~ ability o~ the contra~t
agent, and Adm~nistration route. The optimum concen~ra-
tion for the parama~net~ substance represents a
balance betwee~ various ~a~tors. In general, operatinq
wîth a p~lmary magnet ge~erating a 0.0~ ~ ield,
opt1mum conaentration~ have been ~ound to lie in
the r~nge 1 to lO ~M, espec~ally 3 t~ 9 mM, more
e~pecially ~ to 8 mM and particularly 4.5 to ~.5 mM.
~ompoQitions ~or intravenou6 administra~lon preferably
will contain the parama~netio m~terial at GonaentratiOn
of 10 to lO~0 mM, ~pecially pre~erably 50 to 500 mM.
Por lonic materials the aonc~ntration will pa~tlc~-
larly pre~erably be ~n tha range S0 - 200 ~M, e6pecially
140 to 160 mM and for non-ionic materials 200 -400 mM,
e~pecla~L~ 290 - ~3n mM. For imaging of the urinary
tract or the renal fiyst~m however co~position~ may
perhaps be u~ed having aonaent~tions of for example
10 - 100 mM for ionic or 20 to ~00 mM f~r non-loni~
materials. ~oreove~ ~or bolu~ iA~ect~on, the Concen~ratiOn
may conveniently be 1 t~ lO mM, preferably 3 to g mM
eta.
Thu~ In the ~ethod of the inventloh one may u~e
a COhtraSt medium compri~ing ~ p~y~iologl¢Ally tolQrable
cyclic nitroxide stable free r~dical At a concentra~ion
of from ~ to 500 mM in ~ ~terlle phy~iologi¢ally
tolerable liquid c~rrier, ~ald nitrQxide ha~lng ~n
elea~ron ~pin resonan¢e transition with a llnew~dth
o~ l Gauss or le~s, preferably Less than lO0 mG, at
concen~ratlon~ ~ up to lO mM, especially at 1 or
2 mM.
As mentioned a~ove, th~ first and second radiationx
will ~enerally ~e RF and MW respec~ively and the
radiation sources thus pre~erably are R~ and MW sources.
4l~
- 23 -
The ~irst radiation source is preferably provided
with means ~or ad~usting the pulse timing and duration
so that the desired imaging technique (eg. SR, IR,
SE r FI, etc.) may be chosen and so that the pulse
sequence repetition rate l/TR may be selected to
increase or reduce image acquisition time or to determine
Tl, T2 or nuclear (usually proton) density.
The first radiation source is also preferably
provided with means for adjusting the central frequency,
bandwidth, and intensity of the first radiation pulses.
In MRI, the radiation pulse which excites the
resonating nuclei is applied while the sample is
in a strong magnetic field conventionally with a
field gradient in one direction (e.g. the Z direction~.
The central frequency and bandwidth of the nuclei
exciting pulse together with the Z direction field
gradient during the exciting pulse serve to define
the position along the Z axis and the thickness in
the Z direction of the slice perpendicular to the
Z axis containing nuclei whose spin transitions are
excited by that pulse. Thus, for example, Fourier
transformation of a square wave pulse of central
frequency VO would show such a pulse to contain a
range of frequencies centered about VO and each correspon-
ding to the Larmor frequency of resonating nuclei
in a particular X~ plane along the 2 axis. Thus
by providing the apparatus with means for adjusting
or selecting the central frequency and bandwidth
o the first radiation, the section through the sample
(the image zone) and of course the isotopic nature
and chemical environment of the resonating nuclei
may be selected.
The second radiation source may be a continuous
wave tCW) transmitter or alternatively may be arranged
to emit pulses or trains of pulses of the second
radiation.
~3~ 8
- 24 - 20208-1346
As with the resonating nuclei, the Larmor frequency
of the unpaired electron coupling with the resonating nuclei is
also dependent on the local magnetic field and not only will the
esr transition have a finite linewidth in the esr spectrum, but
that spectrum will generally also show some fine structure, i.e.
splitting due to the fields generating by non-zero spin nuclei in
the paramagnetic material.
To achieve full benefit of the amplified FID signal of
the nuclear spin system and to minimize the dosage of the contrast
agent (if required), it is therefore beneficial to excite and
preferably saturate the electron spin system using a range of
frequencies matched to the frequencies of all or most of the peaks
in the esr spectrum. This can be done by use of a second radiation
source emitting a band of frequencies (e.g. in pulse trains) or by
use of two or more sources emitting at different frequencies.
To achieve the desired frequency spread in the second
radiation, it may be desirable to use pulses of relatively short
duration (hereinafter "micropulses"), for example of the order of
nano or microseconds, and to optimize the amplified population
difference of the nuclear spin system by keeping the esr transi-
tion at or near saturation it may thus be desirable to arrange the
second radiation source to emit a train of micropulses, the
adjacent micropulses being so spaced as not to permit serious
longitudinal relaxation of the electron spin system in the periods
between the micropulses.
~iL248
- 24a - 20208-1346
Alternatively, by providing a decoupling means compris-
ing a third radiation source capable of exciting spin transitions
in certain nuclei (other than the resonating nuclei) the number
of peaks in the esr spectrum or the linewidth of a broad peak
may be reduced. Thus
~3(~:1Z~
- 25 -
multiple peaks in the esr spectrum of the unpaired
electron can arise from coupling betwen the spins
of the electron and nearby non-zero spin nuclei (the
transition splitting nuclei) in the same molecule.
Where the transition splitting nuclei are not the
resonating nuclei for the MRI procedure (for example
where they are of different isotopic nature or, if
they are of the same isotopic nature, where their
chemical shifts are such that their Larmor frequencies
are sufficiently distant from that of the resonating
nuclei in the same region that they are not excited
by the first radiation), the spins of the unpaired
electrons and the transition splitting nuclei can
be decoupled by saturating the nmr transition of
the transition splitting nuclei with a high intensity
radiation at the Larmor frequency of the transition
splitting nuclei (which as indicated above would
not be close to the Larmor frequency of the resonating
nuclei~. With such saturation, the hyperfine structure
in the esr spectrum disappears to leave a single
peak and the esr transition can readily be saturated
using a single second radiation source as previously
discussed. For this mode of operation, the apparatus
of the invention should be provided with means for
emitting the third radiation. The third radiation
emission may be continuous or pulsed (or may take
form of a continuous train on a series of trains
of micropulses as described earlier for the second
radiation) and suitably is emitted over substantially
the same period as the second radiation.
The second radiation source(s) and, where present,
the third radiation source will therefore, like the
first radiation sourcei~preferably be provided with
means for adjusting pulse timing, pulse duration,
central frequency, bandwidth and intensity if they
are pulsed sources, and central frequency, bandwidth
and intensity if they are CW emitters.
~3~3~Z~8
- 26 -
~ he sample may be exposed to the second radiation
either continuously or for one or more periods between
the initiating pulses of subsequent first radiation
pulse sequences. ~referably, exposure to the second
radiation will be in the period in which no field
gradients are imposed on the sample, e.g. for at
least part, and preferably all, of the delay period
between the final FID signal detection period of
each seauence and the initial first radiation pulse
of the next.
Since the invention permits images to be obtained
with adequate resolution at lower than conventional
main magnetic fields, the primary maqnet in the apparatus
of the invention may, if desired, be arranged for
operation at low fields, e.g. 0.002 to 0.1 T, especially
about 0.02 T, or even as low as the ambient magnetic
fieldr i.e. about 0.5 Gauss. Low field operation
is particularly preferred not only for economic reasons
but also to minimize MW heating of the subject and
to improve tissue contrast which is generally found
to increase with decreasing field strength.
The apparatus of the present invention should
particularly preferably be arranged for operation
both with and without the amplified FID in order
that conventional imaging may also be performed on
the same apparatus.
The apparatus of the invention is arranged
to allow MRI of the sample to be performed and in
cetain instances may simply constitute a conventional
MRI apparatus adapted by the provision of a second
radiation source as described above. ~he MRI procedure
involved in the use of the apparatus and the method
of the invention may also involve any one of the
conventional image generation procedures, such as
for example back projection or three- or two-dimensional
13~ 8
Fourier transformation (3DFT and 2DFT~, although
the latter two of these may generally be preferred.
In 2DFT, the sample is placed in a strong magnetic
field (the field direction being the Z direction)
and is allowed to equilibrate. A small field gradient
(the slice selection gradient~ is then applied, e.g~
in the Z direction, and while the slice selection
gradient is superimposed on the main field the sample
is exposed to an RF pulse (the initiating pulse)
of a given central frequency, bandwidth and duration.
Together, the central frequency, the bandwidth and
the combination of the main field and the slice selection
gradient serve to define the position and thickness
of the image zone, the tomographic section through
the sample transverse to the slice selection gradient
in which the resonating nuclei will be excited by
the RF pulse. The duration of the pulse determines
the resultant change in transverse and longitudinal
magnetization of the resonating nuclei. With a 90
pulse, after the slice selection gradient and the
RF pulse are simultaneously terminated, a small field
gradient ~the phase encoding gradient) is then imposed
for a short period in a direction transverse to the
slice selection gradient, e.g. in the Y direction,
causing the phase of the oscillating FID signal to
become dependant on the position in the Y direction
of the signal's source and thus encoding spatial
information in the phase of the FIV signal. After
the phase encoding gradient is terminated, a third
small field gradient (the read gradient) in a direction
perpendicular to the previous two (the X direction)
is imposed to encode spatial information in the FID
frequency and the FID signal is detected and its
intensity as a function of time is recorded during
the imposition of the read gradient.
~L3(3~Z~E~
- 2~ ~
The FID si~nal that is detected is the combination
of signals from resonating nuclei throughout the
imaqe zone. If in simple terms it is viewed as the
sum of signals from an array of sources extending
in the XY plane, the oscillating signal from each
source will have an overall intensity dependent on
the local density of the resonating nuclei, a frequency
dependant on the position of the source in the X
direction and a phase dependant on the position of
the source in the Y direction.
The read gradient is terminated after the FID
signal decays and, after a delay time to permit equilibra-
tion, the slice selection gradient is reimposed and
the initiating RF pulse of the subsequent pulse sequence
is applied.
Image generation requires detections of the
FID signal for a series of pulse sequences, each
with a phase encoding gradient of different strength
or durationr and two-dimensional Fourier transformation
of the resultant data can extract the spatial information
to construct a two dimensional image, in the case
described an SR image.
Different imaging techniques, such as IR, SE,
etc., or different image generation techniques, e.g.
simultaneou~ slice, volume acquisition, back pro~ection
etc., will of course require different pulse and
field gradient imposition sequences, sequences which
are conventional in the art.
An embodiment of the invention will now be
described further by way of example and with reference
to the accompanying drawings, in which:-
- 29 -
Fiqure I i5 a schematic perspective drawing of an
MRI apparatus according to the present invention;
Figure 2 is a schematic perspective drawing of the
emitters of the first and second radiation in the
apparatus of Figure 1.
Referring to Figure 1, there is shown an MRI apparatus
1 having a sample 2, aosed with a paramagnetic contrast
agent according to the invention, placed at the axis
of the coils of electromagnet 3. Power from DC supply
4 to the electromagnet 3 enables the strong main
magnetic field, for example a 200 Gauss field, to
be generated.
The apparatus is further provided with resonators
S and 6 for emitting the first and second radiations
respectively. Resonator 5 is connected to RF transceiver
7 powered by power supply 8 and resonator 6 is connected,
for example by waveguides, to microwave generator
9 which is powered by power supply 10.
Microwave generator 9 may be arranged to emit
MW radiation having more than one maximum frequency
in order to excite more than one esr transition.
The frequency selection, bandwidth, pulse duration
and pulse timing of the flrst and second radiations
emitted by resonators 5 and 6 are controlled by control
computer 11 and interface module 18.
Computer 11 also controls the power supply
from power sources 12, 13 and 14 to the three pairs
of Helmholtz coils 15,-I6 and 17 which are shown
in further detail in Figure 2. The coils of coil
pair 15 are coaxial with the coils of electromagnet
3 and the saddle coils of coil pairs 16 and 17 are
arranged symmetrically about that axis, the Z axis,
~3a~z~
- 30 -
~ith their own axes mutually perpendicular and perpendi-
cular to the z axis. ~oil pairs 15, 16 and 17 are
used to generate the magnetic field gradients that
are superimposed on the main field at various stages
of the imaginq procedure, e.g. in two-dimensional
Fourier transform imaging, and the timing sequence
for operation of the coil pairs and for operation
of the MW generator and the RF transceiver is controlled
by computer 11 and interface module 18.
Where a contrast agent is to be used which
has a multiplet in its esr spectrum, the apparatus
may also be provided with decoupler comprising a
further ~F resonator 19 ~shown with broken lines)
connected to an RF transmitter and a power supply
(not shown) and controlled by computer 11. The decoupler
may be operated to emit a third radiation at a frequency
selected excite the nuclear spin transition in non-
zero spin nuclei in the contrast agent.
In operation in MRI, the power supply to the
electromagnet 3 is switched on and an essentially
uniform main magnetic field is generated within the
cavity within its coils. The magnitude of the main
field generated by electromagnet 3 is maintained
essentially constant throughout the imaging procedure.
The sample 2, for example a patient, is placed
within the coil cavity and after a short delay, for
example several seconds, the imaging procedure can
begin.
Interface module 18 activates the power supply
to coil pair 15 for a short time period during which
DC current flowing through the coils of coil pair
15 in opposite directions about the Z axis results
in an approximately linear field gradient in the
Z direction being imposed on the main field.
- 3~ -
~ ithin the time period for which coil pair
15 is energized, interface module 18 activates RF
transceiver 7 to cause resonator 5 to emit an RF
pulse, e.g. a 90 pulse, to excite the nmr transition
of those resonating nuclei tgenerally protons~ whose
Larmor frequencies correspond to the frequency band
of the RF pulse. The duration, intensity, band width
and central frequency of the RF pulse may be selected
by computer 11. For a given isotope in a given chemical
environment, the ma~or determining factor for the
Larmor fre~uency will be the magnitude of the externally
applied magnetic field, and thus effectively the
RF pulse serves to excite the MR transition of the
selected non-zero nuclear spin isotope (generally
water protons) within a cross-section (the image
zone) of the sample that is transverse to but has
thickness in the Z direction.
On termination of the RF pulse, current in
coil pair 15 is also terminated and after a very
short delay interface module 18 energizes coil pair
16 to provide a field gradient in the Y direction
for a short time period. This is termed the phase
encoding gradient as the field gradient causes the
Larmor fre~uency for the resonating nuclei to vary
linearly across the image zone in the Y direction
for the period that coil pair 15 is energized. With
the remova:L of the perturbation of the Larmor freauencies
on termination of the phase encoding gradient, the
oscillation frequencies of the contributions to the
FI~ signal from different source areas of the image
zone return to being substantially the same, but
the phases of such contributions are shifted to an
extent dependant on the~Location of the particular
source area along the Y direction.
~ fter terminating current in coil pair 16,
the interface module 18 then energizes coil pair
~3-~iJLZ~8
- 32 -
l7 to provide a field gradient (the read gradient)
in the x direction, and reactivates RF transceiver
7 to detect ~he FID signal from the sample.
The FID signal is assumed to arise from the
transverse magnetization of the nuclear spin system
within the image zone since the MR transition was
excited by the RF pulse for resonating nuclei in
this zone only. As described above, the intensity
of the FID signal as a function of time contains
encoded information regarding the distribution of
the resonating nuclei in the imaqe zone in the X
and Y directions respecti~ely .
The FID signal intensity falls off rapidly
and exponentially with time as the system dephases
and the period for which the read gradient is imposed
and the transceiver 7 detects the FID signal from
the sample is generally very short, for example of
the order of milliseconds.
To generate an MR image of the image zone it
is necessary to repeat the pulse and detection sequence
for many further times, e.g. 64-1024 times, each
time generating phase encoding gradients of different
magnitude or duration. Often, to produce a good
S/N ratio, signals for several, e.g. 2-4, identically
performed sequences will be summed. FID signals
for each set of sequences are transformed by the
computer 11 using a standard two-dimensional Fourier
transform algorithm to produce the desired spatial
images of the image zone.
In conventional MRI, after termination of the
only or the last FID signal detection period in a
pulse and detection sequence and before the subsequent
imposition of the slice selection gradient and emission
~3(~ 8
- 33 -
of the initiaing RF pulse of the next sequence, it
has been necessary to wait for a delay period, generally
of the order of seconds, until the resonating nuclei
ha~e relaxed to near equilibrium in order to build
up sufficient longitudinal magnetization for the
FID signal following the new RF pulse to be sufficiently
strong to give an ~cceptable S/N ratio.
However, in the operation of the apparatus
of the present invention, the delay period following
the only or the last detection period may be reduced
by the use of the amplified nuclear population difference
resulting from the coupling between the electron
MR and nuclear MR transitions. Thus at least in the
period between termination of the last read gradient
for each pulse sequence and the emission of the initiating
RF pulse of the next sequence, for example for a
period of about 10 ms to 100 ms, interface module
18 activates MW generator 9 to cause the sample to
be irradiated with MW radiation of a central frequency
corresponding to the Larmor frequency of the paramagnetic
centre in the contrast agent in the sample, either
CW radiation or, preferably, a train of radiation
pulses.
The contrast enhancement in MRI achievable
by the use of paramagnetic contrast agents and of
~W stimulation of esr transitions of the contrast
agents is illustrated by the results set forth in
Table I below.
The table presents contrast enhancement values
for test samples comprising tubes containing solutions
of a range of varying concentrations of a range of stable free
radicals in a range of solvents at a range of different
MW power levels using a primary magnet of 0.02 T.
- 3~ -
The contrast enhancement values are determined
as the ratio of the "areas under the peaks" of the
saturation recovery FID siqnals with and without
imposition of the MW radiation.
The four solvents used were water, Seronorm,
water having reduced dissolved oxygen ("Deoxy-H20")and
Seronorm having reduced dissolved oxygen ("Deoxy-
Seronorm"). Red~ction of dissolved oxygen levels
was achieved by bubblinq nitrogen through water or
Seronorm for about 1 minute. Seronorm is an artificial
human serum available from Nycomed AS, Oslo, Norway.
~31~1248
- 35 - 202~8-1346
Concent- Contrast Enhance-
Compound MW ration ment in
Power (W) (mM) H2O Deoxy H2O Seronorm Deoxy-
Seronorm
o ~ 2 2.5 11 12
96.6 111.3 37.58 40
2 2.5 12 12
80.3 95.1 39.8 43.'
~H 2 2.5 17.5 18.5
~20 5 76.6 94.2 32.7 45.4
~w~
O -
~ 2 2.5 11 14
73.8 98.8 40.7 42.7
o~
1 1.25 19.9 32.9 6.7 10.6
1 2.5 18.9 33.9 9.7 14.7
1 5 17.8 29.6 8.3 12.4
2 1.25 30 S0.0 12 15
2 2.5 30 58 16 26
~3~
.
20208-1346
-- 36 --
Gon~:ent- ~
Compo~tnd MW ration ~ j Contrast ~nhanaemen~ in
Power (w) (mM) ~ H20 I Deoxy-1~2O~ Seronorm ! DeoXy-Seronorm
~_~ 1 2,5 1 21.2 ` 30.1 ~8,6 ~ 12.7
_~ _( o 2D 5 1 93.9 109.Z 1 ~3 ¦ 53~ 8
N~ ~ , !
__ ~
JCûo~
r~ 2 2.5 S8.~
~ 20 5 ~g.8 1, 85.2
O-
N~o~ _ . _ I
~H 2 2. S l 21.6 20. 5
_~_ 2 5 46.3 1 6J.. 1 31.5 29.5
O O ~) _ __ ~ .__
F=~ 2 2.5 12~0 25.~ ¦ 26.0 33.G
44 . 8 65 . 1 l 8~ . 2 96 . 2
F~ -
,L o ~ ~ ~ t - - -- --
?.5 I .... 12 ' 12.1
2 5 43 163.~ 22,7 I~ 22.9
1. ~ .
~I 2.5 79.6 92.1 ~9 2 31 9
_~1,_2 r~
. . .. .. ... .
4E3
- 37 - 20208-1346
Concent- Contrast Enhance-
Compound MW ration ment in
Power (W) (mM) H2O Deoxy-H2O Seronorm Deoxy-
Seronorm
2 5 34 54 15 24
1.25 62.4 73.6 26.4 31.3
2.5 90.3 113.3 46.9 57.4
112.2 139.9 58.2 72.3
o
2 2.5 17 22
57.4 67.9
~JH~,
2 5 48 18
109 55.2
N0 (503~L 2 5 73
~F ~m~'S 20 5 92 90.3
~ompounds are commercially available or are preparable
ethods described in the literature mentioned herein.
~3~1Z48
- 38 -
Thus, to summarize, the present invention opens
up new possi~ilities for ~RI, in particular:-
(i~ It opens the possibility of creating enhancedimage contrast utilizing paramagnetic species naturally
occurring in the tissue. Such con~rast might under
certain conditions be achieved without administration
of a contrast agent to the subject thus avoiding
all associated toxicity and excretion problems and
contrast enhancement of tissue to which it is difficult
to deliver a contrast agent may well be permitted.
(ii) It allows image acquisition times to be reduced
without undue decrease in resolution, even using
FI techniques.
(iii) It allows reduced field strength magnets to
be used without undue decrease in resolution.
(iv) ~t allows the S/N ratio and hence spatial resolu-
tion to be increased without undue increase in image
acquisition times, thereby perhaps allowing features
otherwise obscured by noise to be detected.
(v) It allows highly enhanced image contrast for
tissues capable of accumulating the paramagnetic
material, possibly acheiving contrast effects well
above 1 for such tissues.
(vi) Contrast enhancement may be achieved at lower
contrast agent dosages, or concentrations at tissue
sites of interest, than with conventional contrast
agents.
