Key words
ophthalmologic implant - magnetic resonance imaging - MRI safety - magnetic forces
1. Introduction
As a diagnostic method, magnetic resonance imaging (MRI) is used for imaging soft
tissue structures. During an MRI examination, ferromagnetic materials in the patient’s
body can become a potential danger to the patient. The high magnetic field strength
can result in movement of the implant, causing irreversible damage to sensitive organs
such as the eye [1]. The literature has frequently described the risks and physical interactions of
ferromagnetic objects in MRI [2]
[3]
[4]
[5]
[6]. The purpose of this study is to investigate the behavior of an implant under development
in a 3 T MRI unit. This implant is primarily used to treat glaucoma and is positioned
in a pre-prepared scleral pocket in the eye. It consists of a silicone base body into
which a microvalve flap containing a ferromagnetic plate is cut. There is a small
gap between the valve and the base body through which a constant outflow of aqueous
humor is achieved, thus preventing the development of excessive intraocular pressure.
An external magnet is used to regularly open the valve to prevent attachment of fibroblasts.
Fibrosis and associated closure is often a cause of functional failure of conventional
glaucoma drainage implants [7]
[8].
Due to the widespread clinical use of MRI, it is very important to check the MRI suitability
of new implants. Therefore the study will examine in vitro the magnetic forces acting
on an implant.
2. Materials and Methods
2.1 Material
All tests were performed on a 3 T MRI unit (Philips, Achieva 3.0 T TX). Two prototypes
were examined ex vivo, which were implanted in the eye in the region of the sclera
in the head of a freshly sacrificed rabbit. Both prototypes consist of a silicone
body with a microvalve flap made of chromium-nickel steel with dimensions of 0.5 × 0.5 mm2 and a thickness of 50 µm. The two prototypes differ from one another with respect
to their shape and the number of suture points used to fix the implant in the intraocular
tissue in the area of the sclera. Prototype 1 is circular with a diameter of 4 mm
and was attached at three points in the eye. The silicone body of prototype 2 is rectangular
(3 × 2 mm²) and has four suture points.
The properties of the microvalve flaps were determined in vitro prior to the investigations.
Since a metal plate with a size of 0.5 × 0.5 mm2 can be handled in MRI only with great difficulty, 10 insulated metal plates made
of chromium-nickel steel with a thickness of 50 µm and base areas between 1 × 1 and
8 × 8 mm2 were used in the tests ([Table 1]). The related values for the implant prototypes were extrapolated from the results
of these measurements.
Table 1
Overview of the characteristics of the steel plate.
|
no. of SP
|
side-length [mm2]
|
m [mg]
|
FG [mN]
|
|
1
|
8 × 8
|
23.90
|
0.234
|
|
2
|
7 × 7
|
18.50
|
0.181
|
|
3
|
6 × 6
|
13.90
|
0.136
|
|
4
|
5 × 5
|
9.30
|
0.0912
|
|
5
|
4 × 4
|
5.80
|
0.0569
|
|
6
|
3 × 3
|
3.30
|
0.0324
|
|
7
|
2 × 2
|
1.10
|
0.0108
|
|
8
|
1.5 × 1.5
|
0.70
|
0.00 687
|
|
9
|
1.25 × 1.25
|
0.50
|
0.00 491
|
|
10
|
1 × 1
|
0.40
|
0.00 392
|
|
11[1]
|
0.5 × 0.5
|
0.10
|
0.000 981
|
SP – Steel Plate; m – mass; FG – weight forces.
1 extrapolated values of steel plate 11.
2.2 Methods
2.2.1 Measurement of translational force
The standardized ASTM F 2052 deflection angle test [9] was used to measure the translational force upon the metal plates, which involved
a plate fixed to a non-ferromagnetic holder via a free-swinging string and positioned
on the central axis of the MRI in the area of the maximum induced magnetic force ([Fig. 1a, b]). This region was determined using Kemper’s method [10] at the static magnetic field located 86 cm from the isocenter of the magnet. The
deflection angle β in the direction of the vertical z-line of the magnetic field was
read from the string using a protractor. The translational force FT in z‑direction (magnetic field direction) was calculated for each metal plate based
on each angle of deflection according to
Fig. 1 a Schematic illustration of the translational force apparatus [10]. The deflection angle β is measured to the direction of the magnetic field B0. b Translational force apparatus (wooden) with paper protractor and the suture mounting
made of plastic. The steel plate (marked by a red arrow) and the additional weight
(rubber ring) fixed on a suture.
Thus FG denotes the weight force, m the mass of the metal plate, g the acceleration due to
gravity (9.81 m/s2) and β the deflection angle in relation to the vertical.
The implants were additionally weighted with rubber or plastic loads weighing between
0.052 g and 0.440 g to achieve a deflection angle between 25° and 65°. Each metal
plate was measured twice with two different weights. The mean translational force
was then determined and compared with the weight force FG.
2.2.2 Torque determination
Torque M was determined following an internationally-established standard method [11]
[12]
[13], likewise in a static magnetic field. A laminated protractor was positioned horizontally
in the isocenter of the magnet. Each metal plate was then individually put on the
protractor and placed in positions of 45°, 90°, 135°, 180°, 225°, 270° and 315° to
B0 ([Fig. 2]). Two observers analyzed how each metal plate reoriented itself to the magnetic
field B0. This movement was qualitatively evaluated using a 5-point graduation according to
Sommer et al. [11] ([Table 3]).
Table 2
Overview of the weight forces, translational force and torque of the steel plates
in 3 T MRI.
|
no.of SP
|
FG [mN]
|
FT average [mN]
|
σFT [mN]
|
FT/FG
|
M score
|
|
1
|
0.234
|
2.37
|
0.21
|
10.1
|
4
|
|
2
|
0.181
|
1.72
|
0.32
|
9.47
|
4
|
|
3
|
0.136
|
1.26
|
0.20
|
9.21
|
4
|
|
4
|
0.0912
|
0.722
|
0.04
|
7.92
|
4
|
|
5
|
0.0569
|
0.505
|
0.04
|
8.87
|
4
|
|
6
|
0.0324
|
0.298
|
0.05
|
9.20
|
3
|
|
7
|
0.0108
|
0.0791
|
0.04
|
7.32
|
3
|
|
8
|
0.00 687
|
0.0491
|
0.03
|
7.15
|
3
|
|
9
|
0.00 491
|
0.0354
|
0.02
|
7.21
|
3
|
|
10
|
0.00 392
|
0.0562
|
0.04
|
14.3
|
3
|
|
11[1]
|
0.000 981
|
0.00 981
|
/
|
10
|
/
|
SP – Steel Plates; FG – weight forces; FT – translational force; σFT – standard deviation; M – torque.
1 extrapolated values of steel plate 11.
Fig. 2 Laminated angle scale with a metallic steel plate positioned on 45°. This position
describes the angle between the marker in the middle of the plate towards B0.
Table 3
Qualitative evaluation of torque [11].
|
score 0
|
no torque
|
no movements towards B0
|
|
score 1
|
mild torque
|
the object slightly changes orientation but does not align to B0
|
|
score 2
|
moderate torque
|
the object aligns directly to B0
|
|
score 3
|
strong torque
|
the object shows rapid and strong movement to B0
|
|
score 4
|
very strong torque
|
the object shows very rapid and very forceful alignment to B0
|
2.2.3 Checking the functionality of the magnetic valve flap
In order to demonstrate the functionality of the implant prototypes (see section 2.1),
the opening of the flap of each prototype was visually checked with a neodymium magnet
(magnetic flux density of 0.5 T at 1 mm axial distance, 0.26 T at 5 mm distance) using
a reflected-light microscope. The magnet was held very closely to the magnetic valve
flap (distance approx. 1 mm). Prototype 1 implanted in the eye of a rabbit’s head
was used to check whether unintentional opening of the flap occurs during MRI. The
head was placed in various locations both in the gantry area and in the isocenter
of the MRI. Behavior of the flap was observed using a magnifying glass (focal length
f = 5 cm) and an MRI‑compatible light source.
2.2.4 Checking the holding forces on the fixation sutures of the implant
To test the stability of the fixation sutures, prototype 1 was exposed to a force
of 10 mN in the eye of a freshly sacrificed rabbit using a dynamometer. A suture (Vicryl
Plus, 3 – 0) was placed centrally below the implant. Using the suture and dynamometer,
the implant was laterally pulled at an angle of 0° (directly from above), laterally
at an angle of 45° and nasally at an angle of 45°. The implant was subjected to the
force for 20 minutes in each position. Subsequently the retaining sutures were examined
and assessed using a reflected light microscope.
3. Results
3.1 Translational force and torque
The translational force FT calculated for the 10 metal plates (MP) lay between 0.0354 mN (MP 9) and 2.37 mN
(MP 1) ([Table 2]). The translational force increased with the size of the plates and was usually
almost 10 times as great as the weight force FG ([Fig. 3]). When investigating the torques, the square plates, regardless of their size, showed
an orientation with an outer edge parallel to the magnetic field B0 ([Table 2]). Plates 1 to 5 showed an immediate and rapid orientation with respect to B0 (Score 4). Plates 6 to 10 showed a less rapid movement in their preferred direction
compared to plates 1 to 5 (Score 3).
Fig. 3 Translation force as a function of the weight force of the 10 steel plates in a 3 T
MRI (gradient of the straight lines: 9.62; coefficient of determination R2: 0.99).
3.2 Functionality of the magnetic valve flap
The flaps of prototype 1 and prototype 2 opened when the neodymium magnet approached
the implant (distance magnet to implant: about 1 mm). When the influence of the prototype
1 flap on the MRI was checked, the flap did not open in the area of the highest field
gradient (≤ 17 T/m [14]) or in the isocenter.
3.3 Implant fixation suture holding forces
During the subsequent microscopic examination, no changes were visible at the fixation
points or on the scleral tissue.
4. Discussion
This study investigated a new type of magnetic ophthalmological implant regarding
its MRI safety with respect to magnetic translational forces and torque. The strength
of the forces or torques depends on the magnetic field gradient, possibly on the magnetizability
(= magnetic susceptibility χ) of the material, on the magnetic field strength B0, on the mass and position of the implant in the magnetic field as well as on the
geometry of the implant [15]. Ferromagnetic metals such as iron, cobalt and nickel are characterized by a high
susceptibility (e. g. iron χFe ≈ 105) and are therefore considered to be “MRI unsafe” since they are exposed to large
forces in the magnetic field. Likewise, their alloys and many steel grades can also
initially be classified as unsafe without precise knowledge of their susceptibility
[16].
4.1 Translational force
The maximum translational force is achieved where the product of magnetization and
field gradient reaches its maximum. This location is in the region of the gantry opening,
since on the one hand the field strength is so high that the material already shows
a saturated magnetization, while on the other hand there is a large field gradient
due to the divergence of the field lines, and the field gradient goes towards zero.
The translational force therefore increases with approach to the gantry opening, reaches
a maximum in the area of the opening and disappears within the MRI, as a homogeneous
magnetic field is present there. Consequently the patient is exposed to the greatest
translational force when passing through the magnet opening [10]. For the investigated metal plates of different sizes, there was a linear relationship
between their weight and their translational force. This also demonstrates that all
plates have the same composition of chromium and nickel. Each metal plate is subjected
to a translational force that is about 10 times greater than its own weight. The deflection
angle test used is an established method for determining translational force [2]
[17]
[18]
[19]. Without additional weight, the metal plates exhibited a deflection angle of approx.
90°. At a deflection angle of 90°, no calculation of the translational force is possible
due to the tangent in equation (1). Therefore, the metal plates were weighted with
non-magnetic weights to achieve a deflection angle of less than 65° [10]
[20]
[21]. In order to improve the measuring accuracy, the measurements were carried out twice
with different weights [22]. According to ASTM guidelines, objects with a deflection angle of more than 45°
are classified as “MRI unsafe” [9]
[10]. The reason for this is that with these objects, the force acting by the static
magnetic field is greater than the gravitational force acting on the object [10]
[22]. When evaluating this statement, it must be taken into account that the implant
will later be firmly fixed in the eye by sutures. According to Mühlenweg et al. the
risk of dislocation decreases with the age of the implantation (> 6 months) due to
scarring [16]. Even after microsurgical treatment of the sclera, there is a scarring reaction
after a few months [23]. Since a very high translational force acts on the implants which is significantly
higher than their own weight force, the test described in section 2.2.4 should be
reviewed against a worst-case scenario to determine whether the retaining sutures
and silicone can counteract the translational force. The metal plate of prototype
1 weighs 0.1 mg. Interpolation of the values in [Table 2] yields a translational force of 0.01 mN. The dynamometer used pulled on the implant
with a force of 10 mN from different angles for 20 minutes each. On the whole the
sutures were exposed to a thousand times the calculated force. Since the subsequent
microscopic examination revealed no changes in the position of the sutures, it can
be presumed that the implant is not dislocated during the MRI examination, and that
both the sutures and the tissue resist force. However, this method does not permit
a clear statement whether tractive force at the histological level can cause minor
damage or induce inflammation.
4.2 Torque
The strongest torques are to be expected in the isocenter of the magnetic field of
an MRI where the magnetic field is most homogeneous and imaging occurs [10]. Depending on the size of the metal plates, the torque score was between 3 and 4
([Table 2]), i. e. the diagonals of the plates were oriented with a fast and instantaneous
movement parallel to B0. The torque score decreased as the dimensions of the plates decreased ([Table 2]). One reason for this are the lower magnetic moments associated with the smaller
dimensions of the plates [16]. The frictional forces between plate and the test surface did not decrease in the
same way, so that the frictional forces had a stronger influence on smaller plates,
thus reducing the resulting torque [11]. Determination of the torque was methodologically very difficult since there is
no uniform method for quantifying the torque for very small objects. Consequently,
only a qualitative assessment of torque was performed by two independent observers
using a 5-point graduation, which was developed specifically for small objects [10]
[11]
[12]
[13]
[24]
[25]. In contrast to translational force, it is difficult to define an upper safety limit
value for torque [11]. Whereas translational force increases linearly with the field strength, torque
increases in proportion to the square of the field strength [10], and is therefore a considerable and not specifically calculable safety risk [11]. The torque acting on the implant depends not only on the dimensions and susceptibility
of the material but above all on its geometric shape. In particular, elongated objects
are exposed to strong torque, while for square-shaped objects, the torque is usually
lower [10]. Although the metal plates in the isocenter of the MRI are exposed to great torque,
it is not sufficient to cause opening of the valve flap or displacement of the silicone
body.
4.3 Functionality
In addition, according to chapter 2.2.3, it was examined whether the magnetic valve
flap opens in the MRI or whether its function is restricted. When the implant was
positioned in the area of the isocenter and the gantry area, opening of the flap was
not visible using a magnifying glass. In contrast, the flaps of both prototype 1 and
prototype 2 could be opened with a bar magnet, although it has a smaller magnetic
field than the MRI. This can be explained by the fact that the small spatial expanse
of the magnet is associated with a large field gradient, while the field gradient
is lower due to the extended magnetic field of the MRI. In summary, it can be assumed
that the function of the magnetic valve flap is not restricted or disturbed by the
translational forces and torques generated by the magnetic field of the MRT.
5. Conclusions
Due to the small size of the implant only small forces act on it, which can be easily
compensated for by its silicone sheath and suture fixation. An estimation of possible
heating is still necessary for a fundamental assessment of the MRI suitability of
the implant. This was carried out in a second study together with an investigation
of artifact formation.
Clinical Relevance of the Study
-
Due to the frequent clinical use of MRI, new implants must be tested for their MRI
safety.
-
Magnetic forces have been precisely evaluated in order to assess the effect of tractive
forces.
-
The eye, in particular, contains sensitive structures which could incur irreparable
damage through dislocation of the implant.
-
The function of the magnetic valve flap is not restricted or damaged by the magnetic
forces during an MRI examination.
