MRI-INDUCED SAR ON PACEMAKER LEADS

Numerical Simulations on Three Human Phantoms

Eugenio Mattei, Giovanni Calcagnini, Michele Triventi, Federica Censi and Pietro Bartolini

Italian National Institute of Health, Department of Technology and Health, Rome, Italy

Keywords: Magnetic Resonance Imaging, Pacemaker, Specific Absorption Rate, Human Visible Dataset.

Abstract: Numerical simulations were performed to evaluate the Specific Absorption Rate (SAR) induced at the tip of

a pacemaker (PM) implant by the 64 MHz radiofrequency (RF) field used in 1.5T Magnetic Resonance

Imaging (MRI) procedures. The analysis was performed by using a commercial FDTD software (SEMCAD

X, SPEAG, Switzerland) and aimed at the evaluation of the impact that the patient ‘s morphology has on the

induced local SAR at the implant tip. In particular three human phantoms were studied: a 34-year old man

model, a 26-year old woman, and a 6-year old boy. The three phantoms reproduce more than 70 tissues of

the human body with a resolution of 1 mm. Inside each phantoms, realistic implant configurations were

modelled, considering both left and right pectoral implants, and atrial and ventricular stimulations. The local

SAR values at the lead tip was compared for the three phantoms and sensible differences were observed:

with a RF excitation set to produce an whole-body average SAR of 2 W/kg without any implants, local SAR

values ranged from 641W/kg (woman model – right ventricular implant) to 3 W/kg (boy model – left atrium

implant). We also observed that, in general, ventricular implants showed a higher SAR compared to atrial

ones, as well as right pectoral implants compared to left ones. However, not always a higher implant area or

a longer lead path implied higher SAR at the tip, indicating the coupling mechanisms between the implant

and the RF field are likely to be more complex that the only area-dependent induction law.

1 INTRODUCTION

The number of Magnetic Resonance Imaging (MRI)

scans performed annually has increased dramatically

over the past few years. Parallel to the growth and

evolution of the MRI field, is the burgeoning

number of patients benefiting from implantable

cardiac systems including pacemaker (PM) and

implantable cardioverter/defibrillators (ICDs). The

combination of these two growing phenomena

results in an estimated 50-75% probability of a

patient being indicated for an MRI study over the

lifetime of their device; it has created an estimated

200,000 implanted patients who were denied the

MRI scan, and this numbers are likely to increase in

the future (Rougin et al, 2004). Given the rapid

expansion of technology in the fields of both MRI

and device arrhythmia management, there is

increasing interest in the issue of implantable device

safety in the MRI environment. For the purpose of

MRI, non-ferromagnetic material is available for

manufacturing of implantable devices. Considering

the impressive progress in the use of diamagnetic

material, the most important safety problem

associated with MRI and medical implantable

devices is the potential tissue heating induced by the

radiofrequency (RF) fields. In this filed, numerical

studies are crucial to extend the range of

experimental measurements and to correlate heating

results to those expected in humans. Simulations can

be used to model realistic patient geometries, to

deliver more information (e.g., 3D fields instead of

single measurement points) and to study individually

the impact of parameters such as tissue properties,

boundary conditions, etc. As the finite difference

time domain (FDTD) method has been a widely

used technique for characterization of RF heating, it

is an excellent candidate to render an accurate

estimate of the specific absorption rate (SAR), and

therefore heating, due to implantable devices during

MRI experiments.

The anatomical structure of the human body as

sets of minute elements (voxels) suitable to be

imported inside the FDTD environment have been

obtained from MRI scans, X-ray computed

tomography, or anatomical coloured images of the

135

Mattei E., Calcagnini G., Triventi M., censi F. and Bartolini P..

MRI-INDUCED SAR ON PACEMAKER LEADS - Numerical Simulations on Three Human Phantoms.

DOI: 10.5220/0003131501350139

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2011), pages 135-139

ISBN: 978-989-8425-37-9

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

Visible Human Project (VHP). These data are today

largely available at a high resolution, so that almost

all the various human tissues can be taken into

account (Visible Human Dataset – VHD).

The purpose of this study is to numerically

compute the local SAR induced by a 64 MHz RF

coil used in MRI procedures at the lead tip of a PM

implant. Local SAR values were compared for three

human phantoms: a 34-year-old man model, a 26-

year-old woman, and a 6-year-old boy. Inside each

phantom, realistic implant configurations were

modelled; left and right pectoral implants, as well as

atrial and ventricular stimulations were taken into

account.

2 METHODS AND MATERIALS

The MRI RF coil model was developed using a

commercial FDTD (Finite Difference Time Domain)

solver (SEMCAD X, SPEAG, Zurich, Switzerland).

It represents a 16-rung low-pass birdcage coil (60

cm high, with an inner radius of 30 cm) which

reproduces the RF field generated during MRI

procedures (Figure 1).

Figure 1: FDTD model of the RF birdcage coil.

Two voltage sources were applied at one of the

two external rings of the coil, with a 90° shift both in

space and in time. Inside the coil, a uniform and

circularly polarized magnetic field was thus

generated. An external metal shield was used to limit

the RF field inside the birdcage coil.

The human phantoms were imported from “The

Virtual Family” package developed by the

Foundation for Research on Information Technology

in Society (IT’IS Foundation - Zurich, Switzerland ).

In particular we used:

• The 34-year-old man model;

• The 26-year-old woman;

• The 6-year-old boy model.

Each phantoms distinguishes more than 70 tissues

with a spatial resolution of 1 mm. The phantoms

were placed inside the RF coil and their position was

adjusted to have the coil isocenter at the lowest part

of the sternum (xiphoid process – Figure 2).

Figure 2: Human phantom (female) inside the birdcage

coil.

As PM implant, a bicameral stimulator was

modelled in the right and left pectoral regions. The

lead path was derived from RX images of patients

with a PM implant (Figure 3). Both for atrial and

ventricular stimulation, a unipolar lead was

modelled as a perfect electric conductor (PEC) thin

wire (radius = 0.5 mm) with a silicon insulation of

1.5 mm radius. At the end of the lead, a 1 mm bear

PEC tip was put in contact with the heart wall.

Figure 3: Human phantoms and PM implants.

BIODEVICES 2011 - International Conference on Biomedical Electronics and Devices

136

To obtain the required resolution at the implant,

a two-step approach has been employed: a first

simulation of the birdcage without the wire is

performed using a relatively coarse grid (≈15 million

cells, graded mesh). The fields are recorded on the

surface of a rectangular box around the implant, and

are used to excite a second simulation restricted to

the area where the PM and its lead are implanted

(≈20-30 million cells, graded mesh).

This two-step approach, which is based on

Huygens’ principle, allows to obtain a high

resolution around the implant and to properly model

the lead. To ensure the continuity of the inner

conductor of the lead and of its insulation a maximal

refinement at the lead (0.2-0.6 mm) was needed. The

same mesh parameters were used for the three

phantoms. In Figure 4 the graded mesh adopted in

the area of the implant and the resulting voxel are

overlaid to the hart and vessel of the VHD.

Figure 4: FDTD graded mesh in the area of the PM

implant: right pectoral implant (a) and left pectoral

implant (b).

Local SAR deposition was calculated from the

E-field estimated by the model. SAR was calculated

as described in the IEEE1528 standard (IEEE1528,

2006), over a 1mg mass.

A 1 mg mass was chosen as a trade-off between

a volume small enough to significantly account for

local SAR value, but big enough to prevent

misleading results due to computational errors.

3 RESULTS

Figure 5 shows the SAR distribution for the female

phantom with the PM in the right pectoral region,

exposed to the RF field of the birdcage coil.

Local hot-spots can be observed at the contact points

between the two lead tips and the heart wall, in the

atrium and in the ventricle. A similar distribution

was obtained also for the man and the boy

phantoms.

The local SAR values computed for the man, the

woman and the boy phantom are reported in Figure

6. The highest SAR was observed at the tip of the

right ventricular implant in the woman model (641

W/kg); on the other hand, the boy phantom is the

one associated to the lowest SAR values, for all the

implant configurations tested.

In general, the right pectoral positioning implies

higher SAR compared to left pectoral implants, as

well as ventricular stimulation compare to atrial one.

The only exception is represented by the left

atrial implant in the woman phantom, which showed

higher SAR values than the right atrial one.

In these simulations, the excitation signal applied to

the RF coil was not the same for the man, the

woman and the boy model, but was adjusted to

obtain an average whole-body SAR of 2 W/kg

(maximum value allowed during standard MRI

procedures) inside all the phantoms.

Figure 5: Example of the SAR distribution over the heart

surface resulting from the exposure of the implanted

phantom to the RF field generated by the birdcage coil.

The two hot spots represent the contact points between the

lead tips and the heart wall.

)

MRI-INDUCED SAR ON PACEMAKER LEADS - Numerical Simulations on Three Human Phantoms

137

Figure 6: Local SAR values at the PM lead tip for the

man, woman and boy phantoms. SAR calculation was

averaged over 1 mg mass. The whole body SAR without

the implant was 2 W/kg in all the phantoms.

If the amplitude of the signal is not changed, the

average SAR significantly differs in the three

models: an excitation that produces an average

whole body SAR of 2 W/kg in the boy model, for

example, leads to an average SAR of 38.5 W/kg in

the man phantom, and of 25.9 in the woman

phantom.

In order to characterize the lead paths that a

realistic PM implant implies in different anatomical

structures and in different locations, we calculated

the lead length from the connection with the PM

chassis to the tip. The data reported in table 1 show

how a longer lead does not always imply a higher

SAR at the tip.

Table 1: Lead length for the configuration tested inside the

man, the woman and the boy phantoms.

Phantom type

Lead length (mm)

Atrium Ventricle

MAN

Left imp. 280 365

Right imp. 302 381

WOMAN

Left imp. 283 345

Right imp. 308 369

BOY

Left imp. 148 205

Right imp. 161 207

4 DISCUSSION

RF induced heating of biological tissue has long

been a concern for patients undergoing magnetic

resonance imaging (MRI). With regards to the MRI

induced heating on PM and ICD leads, a wide

database of experimental measurements is now

available in literature (Mattei et al, 2008; Nordbeck

et al, 2009). However, these studies use simplified

model of the human body, typically a rectangular

box phantom filled with an homogeneous gel, whose

physical properties are defined to closely match

those of biological tissues (ASTM F2182–02a).

Even when phantoms of more complex shapes are

used, they always assume a uniform behaviour of

human tissues and do not take into account the

realistic anatomical structure of the human body. In

addition, the large number of variables that take part

in the process may often result in a loss a general

validity, requiring additional efforts to perform

extensive and exhaustive measurements. Thus,

modellistic approaches based on numerical tools

might represent a useful mean, able to overcome

such limitations.

In this study we compared the RF-induced SAR

during MRI scans at the lead tip of a PM implanted

in human models that thoroughly reproduce the

anatomical structures of an adult male, an adult

woman and a boy. Realistic implant configurations

markedly differ for the three phantoms, in terms of

lead length, path, and area covered by the implant.

Thus, it is not surprising that also the local SAR

induced at the lead tip sensibly varies in the

simulations we performed. However, there are

several aspects that suggests how the coupling

mechanisms between the RF field and the PM

implant is much more complex than it may appear.

Several papers in the literature (Sommer et al,

2000, Rezai et al, 2005) chose a configuration of the

pacemaker lead in the coronal plane to achieve a

maximal magnetic induction area in order to

maximize the heating at the lead tip. We found that

the induced SAR is not always proportional to this

area. In particular, in the comparison between left

and right implant configurations, right implants

covered an area significantly smaller than the left

counterparts and the SAR is generally higher for the

former than the latter ones.

The lead length seems to be a parameter that

better correlates with the induced SAR (right

implant implies longer leads than left ones), but also

in this case, simulated data highlight some

exceptions: in particular, the highest SAR was

observed at the tip of the right ventricular implant in

BIODEVICES 2011 - International Conference on Biomedical Electronics and Devices

138

the woman model, which has a shorter lead length

than the counterpart in the man model. Thus, there

should be other aspects related to the morphology of

the human phantom that can affect the amount of the

deposited power and consequently the tissue heating

at the lead tip; such aspects may involve the

particular path the lead follows inside the human

tissues and the electromagnetic field distribution

induced during MRI scans in the same regions.

5 CONCLUSIONS

The present numerical study shows how the

differences in terms of patient’s anatomy (different

gender and age) have an impact on the MRI induced

SAR, and consequently tissue heating, at the tip of

an implanted PM lead. It is justified by the changing

in lead path and electromagnetic field distribution

that different anatomical models imply. Our data

show also that the implant area or lead length are not

the only parameters that can affect the amount of

induced heating the implant tip, but more complex

coupling phenomena must be taken into account.

REFERENCES

ASTM F2182–02a – ‘‘Standard Test Method for

Measurement of Radio Frequency Induced Heating

Near Passive Implants During Magnetic Resonance

Imaging’’, ASTM International, 100 Barr Harbor

Drive, PO Box C700, West Conshohocken, PA,

19428–2959 USA

IEEE, IEEE1528, recommended practice for determining

the spatial-peak specific absorption rate (SAR) in the

human body due to wireless communications devices:

Measurement techniques, IEEE Standards Department

52(5), 1 2006

Mattei E, Triventi M, Calcagnini G, Censi F, Kainz W,

Mendoza G, Bassen HI, Bartolini P. Complexity of

MRI induced heating on metallic leads: experimental

measurements of 374 configurations Biomed Eng

Online. 2008 Mar 3;7:11

Nordbeck P, Weiss I, Ehses P, Ritter O, Warmuth M,

Fidler F, Herold V, Jakob PM, Ladd ME, Quick HH,

Bauer WR. Measuring RF-induced currents inside

implants: Impact of device configuration on MRI

safety of cardiac pacemaker leads. Magn Reson Med.

2009 Mar;61(3):570-8.

Rezai AR, Baker KB, Tkach JA, Phillips M, Hrdlicka G,

Sharan AD, Nyenhuis J, Ruggieri P, Shellock FG,

Henderson J. Is magnetic resonance imaging safe for

patients with neurostimulation systems used for deep

brain stimulation? Neurosurgery 2005

Nov;57(5):1056-62

Roguin A, Zviman MM, Meininger GR, Rodrigues ER,

Dickfeld TM, Bluemke DA, Lardo A, Berger RD,

Calkins H, Halperin HR. Modern pacemaker and

implantable cardioverter/defibrillator systems can be

magnetic resonance imaging safe: in vitro and in vivo

assessment of safety and function at 1.5 T.

Circulation. 2004 Aug 3;110(5):475-82.

Sommer T, Vahlhaus C, Lauck G, von Smekal A, Reinke

M, Hofer U, Block W, Traber F, Schneider C, Gieseke

J, Jung W, Schild H. MR imaging and cardiac

pacemakers: in-vitro evaluation and in-vivo studies in

51 patients at 0.5 T. Radiology 2000 Jun;215(3):869-

MRI-INDUCED SAR ON PACEMAKER LEADS - Numerical Simulations on Three Human Phantoms

139