Magnetic Resonance Imaging at 7 Tesla with Dedicated

Radiofrequency Coils

Application to Cervical Cord and Knee

Maria Evelina Fantacci

1,2

, Laura Biagi

, Mirco Cosottini

4,5

, Mauro Costagli

3,5

, Massimo Marletta

Alessandra Retico

, Riccardo Stara

1,2,7

, Mark Symms

, Gianluigi Tiberi

3,5

, Virna Zampa

and Michela Tosetti

3,5

Dipartimento di Fisica, Università di Pisa, Largo Pontecorvo 3, Pisa, Italy

Istituto Nazionale di Fisica Nucleare (INFN, Sez. Pisa), Pisa, Italy

IRCCS Stella Maris, Pisa, Calambrone, Pisa, Italy

Dip. di Ricerca Traslazionale e delle Nuove Tecnologie in Medicina e Chirurgia, Univ. di Pisa, Pisa, Italy

Fondazione IMAGO7, Pisa, Italy

Dipartimento di Radiologia Diagnostica ed Interventistica AOUP, Pisa, Italy

Lucas center for Imaging, Department of Radiology, Stanford University, Stanford, CA 94305, U.S.A.

General Electric ASL (EMEA), Pisa, Italy

Keywords: Ultra High Field Magnetic Resonance Imaging Coils, Cervical Cord UHF MRI, Knee UHF MRI.

Abstract: Magnetic Resonance (MR) Imaging is a valuable tool in the diagnosis and monitoring of various

musculoskeletal pathologies. New Ultra-High Field (UHF) 7 T MRI systems, with their enhanced Signal-to-

Noise Ratio, may offer increased image quality in terms of spatial resolution and/or shorter scanning time

compared to lower field systems. However, these benefits can be difficult to obtain because of increased

radio-frequency (RF) inhomogeneity, increased Specific Absorption Rate (SAR) and the relative lack of

specialized and commercially available RF coils compared to lower field systems. This study reports the

feasibility of imaging in bones and cartilages at UHF with a 7 T MR scanner available at the IMAGO7

Foundation (Pisa, Italy). Dedicated radio-frequency coils for proton imaging have been designed,

developed, optimized for different anatomical regions and validated in vivo, and are now ready for clinical

research studies. The performance of the RF coil prototypes in targeting different anatomical regions are

also demonstrated, obtaining images of the neck (the cervical cord) and of the knee (trabecular bone and

cartilages).

1 INTRODUCTION

The current research in the field of Magnetic

Resonance (MR) biomedical imaging (MRI) is

moving towards increasingly higher static magnetic

field strengths. Whereas 3 Tesla scanners are

becoming widespread in clinical applications,

scanners working at higher static magnetic fields are

currently available only at a limited number of

laboratories in the world, and only for research

purposes.

There are about 60 MR systems for human

studies operating at 7 Tesla or above worldwide, and

they have already demonstrated the great capability

and potential of Ultra-High Field (UHF) MR, and

many technical challenges remain (Ugurbil, 2003;

Kraff, 2015).

The IMAGO7 Foundation in Pisa (Italy) owns

and manages the first and only 7 Tesla whole-body

MR scanner (950-MR scanner, GE Medical

Systems, Milwaukee, WI) in Italy. In this

framework, a research collaboration between the

IMAGO7 Foundation and the Italian National

Institute for Nuclear Physics (INFN) aims to develop

important hardware components, such as RF coils

for specific MR applications, and to exploit the UHF

Fantacci, M., Biagi, L., Cosottini, M., Costagli, M., Marletta, M., Retico, A., Stara, R., Symms, M., Tiberi, G., Zampa, V. and Tosetti, M.

Magnetic Resonance Imaging at 7 Tesla with Dedicated Radiofrequency Coils - Application to Cervical Cord and Knee.

DOI: 10.5220/0005774102290234

In Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2016) - Volume 1: BIODEVICES, pages 229-234

ISBN: 978-989-758-170-0

229

potential in several research areas, including MSK

imaging.

Among the clinical applications that will benefit

from the improved resolution and signal-to-noise

ratio (SNR) obtainable at high magnetic field is the

muscoloskeletal (MSK) system. UHF MR imaging

of the MSK system, including small joints, offers

important potential advantages over lower field

systems in increased sensitivity and enhanced

contrast. However, these benefits can be difficult to

obtain because of increased radio-frequency (RF)

inhomogeneity, increased Specific Absorption Rate

(SAR) and the relative lack of specialized and

commercially available RF coils compared to lower

field systems.

A number of in-vivo studies of the

musculoskeletal (MSK) system at high field

strengths have already been carried out (Majundar,

2008; Farooki, 2002; Gambarota, 2007; Regatte,

2007; Krug, 2009), including the investigation of the

possibility to perform sequential studies during the

clinical and instrumental follow up of

neuromuscular disorders (Retico, 2015). The latter

can be useful for a variety of applications: a) to

allow an earlier diagnosis also in asymptomatic

patients; b) to improve the monitoring of the

progression of muscle involvement; c) to provide

valuable information on the efficacy of ongoing

therapeutic studies (drugs, gene or stem cell

therapy), representing a possible alternative to serial

muscle biopsies. Moreover, at UHF, for this

application fat suppression should be used to reduce

the chemical shift artifacts between fat and water

frequencies.

This paper presents two RF coil prototypes

suitable for 7T MSK applications targeting the

geometry of different anatomical regions (neck and

knee) based on quadrature Tx/Rx surface RF coil

suitable for the detection of the proton signal and the

first images acquired in vivo by means of these

dedicated RF coil prototypes. In fact, dedicated neck

coils are fundamental for studying the spinal cord in

several pathologies of the central nervous system

such as multiple sclerosis, or myelopathies of

different origin. In particular the high resolution and

SNR of the UHF can be exploited to investigate the

gray matter and the fiber bundles within the spinal

cord of patients with amyotrophic lateral sclerosis.

Moreover, UHF MRI of the knee (Krug, 2009) can

allow an accurate characterization of morphology

and biochemical quality of the cartilages for clinical

assessment of early pathological conditions of

cartilage in osteoarthritis, and can allow the

quantification of trabecular bone architecture useful

for clinical assessment of osteoporosis.

This paper is organized as follows: first, the

choice of the coil design is motivated according to

the available MR system and the anatomical region

under investigation; then, the choice of the hardware

components and the coil construction details are

provided; finally the results, consisting of the first in

vivo images acquired on human subjects, are

presented.

2 MATERIALS AND METHODS

2.1 Quadrature Tx/Rx Surface RF

Coils for 7 T MRI

The availability of commercial RF coils for UHF

MR systems is still limited. Therefore, UHF research

sites often set up RF laboratories to develop suitable

coils for specific applications.

The choice of the RF coil design depends on the

target anatomy, on the desired MR acquisition

modality (e.g. MRI, MR spectroscopy (MRS),

multinuclear MRS, …) and it is constrained by the

available MR system equipment.

This study focuses on the adult human neck (for

the cervical cord applications) and knee (for

trabecular bone and cartilage applications). Two

coils have been designed, with geometrical

optimization for accommodating an adult human

neck and an adult human knee, respectively.

The MR system available at the IMAGO7

Foundation is a 7 T scanner currently equipped with

two channels for transmission, which can be either

both for proton (i.e.

H, I and Q channels) or one for

proton and one for another nucleus (i.e.

H and MNS

channel).

As a linearly polarized surface coil for proton

MRI is expected to suffer transmit field

inhomogeneity problems typical at UHF, in an

attempt to obtain more informative structural images

a specific coil design has been used. It consists of a

quadrature surface coil, where the two channels are

both used for proton.

For both the prototypes the coil housing is

designed using the AutoCAD CAD package; the

supports have to be designed according to the target

anatomy, and has to be comfortable, easy to use and

safe. To this purpose we left 10 mm between the

inner surface (where the coil circuits will lie) and the

outer layer (where the neck or the knee will be

positioned). Concerning the choice of the materials,

we used for the mechanical support a polycarbonate

structure obtained by 3D printing.

BIODEVICES 2016 - 9th International Conference on Biomedical Electronics and Devices

230

The coil circuits consist of a printed circuit board

(PCB) in FR4, board thickness 200 micron, copper

thickness 35 micron. The circuits are covered by a

protective paint to avoid oxidation.

The coils presented here are constituted of two

partially overlapped loops driven in quadrature. The

shape of the mechanical frame of the RF coil is a

semi-cylindrical saddle of inner radius 70 mm and

outer radius 85 mm.

The support of the neck coil presents a flat

portion to best accommodate the neck lying on the

coil in the supine position while the support of the

knee coil presents a hemi-cylindrical geometry

(Figure 1).

The RF circuit consists of two loops with

dimensions 170 mm along the horizontal axis, 60

mm and 130 mm along the vertical axial for the

central and peripheral part, respectively. The two

loops are geometrically decoupled by a partial

overlap of 18 mm (adjustable). Referring to Figure

2, the list of components used for assembling the

coil is given in Table 1.

Figure 1: The quadrature Tx/Rx surface RF coil suitable

for the detection of proton signal in the neck (left) and in

the knee (right).

Figure 2: Circuit of the quadrature

H RF surface coil

made by two square loops with partial overlapping.

Table 1: RF components of the single-tuned quadrature

RF coil.

6.8 pF

2,5

5.1 pF

3,4,6,7

4.2 pF

CMS 22 pF

CMP 1.0 pF

The tuning procedure of the

H loop is performed

after measuring the inductance, which implies the

following operation: 1) a known capacitor has been

added to the loop; 2) the correspondent resonance

frequency has been measured through a Vector

Network Analyzer (VNA, E5071C, Agilent

Technologies); the inductance is determined by

analytical calculation. Next, the loop has been tuned

to 298.03 MHz, i.e. the Larmor frequency of the

at 7 T. Next, the coil has to be matched: matching

can be achieved either with or without a load, i.e. in

the unloaded/loaded condition. A capacitive

matching with load is performed; the Q factor is

calculated through appropriate VNA measurements,

and the corresponding resistance is derived

(Mispelter, 2006). The matching capacitor was

determined by using a Smith Chart procedure

(Smith, 1995). The workbench measurements

provided the following values: matching < -17 dB

on both channels, coupling < -15 dB. The Q factors

are 120 and 20 for both channels, in loaded and

unloaded conditions, respectively. The simulated B

and E maps obtained for unitary input power of 1

kW are shown in Figure 3, which demonstrates that

quadrature operation of the 7 T surface coil

improves B

field homogeneity with respect to a

linearly polarized surface coil.

Figure 3: Simulated B

(a) and E (b) field maps of the

quadrature

H RF surface coil for unitary input power of

1kW. Measured B

Magnetic Resonance Imaging at 7 Tesla with Dedicated Radiofrequency Coils - Application to Cervical Cord and Knee

231

2.2 Human Image Acquisitions

Six healthy and one pathological (alteration of the

patellar cartilage) volunteers were considered for the

preliminary acquisitions reported here. Healthy

volunteer age ranged between 24 to 61 years, the

pathological volunteer was 62 years old. In our

Institute clinical studies follow the ethical guidelines

of our local ethics committee. Informed written

parental consent was obtained before enrollment in

the study. As expected, any side effect from muscle

MRI examination has not observed.

Morphological images have been acquired by

means of 3D MERGE and 3D FIESTA sequences,

optimized taking into account the relaxation times of

the tissues of interest at 7 Tesla. In order to evaluate

also the biochemical behaviour of the cartilage in the

pathological subject, T2 and T2* maps (Krug, 2009)

have been also computed.

3 RESULTS

The neck RF coil was used to obtain images of the

neck and the cervical cord in healthy subjects

(Figure 4).

The sequence used to obtain the image reported

in Fig. 4 was a 3D MERGE with 0.5 mm in-plane

resolution, TR = 30 ms, TE = 17.5 ms, thickness =

2.2 mm.

The image obtained demonstrates that this coil

provides excellent anatomical images of the cervical

spine, where spinal gray matter and white matter can

be clearly depicted.

Figure 4: Cervical cord of a healthy volunteer acquired at

7 T with the quadrature

H surface RF coil and a 3D

MERGE sequence. Note the high quality of the image

with a clear depiction of the H shape of the spinal grey

matter.

The knee RF coil was used to obtain images of

the knee in healthy volunteers, in order to assess

both the architecture of the trabecular bone and the

morphology of the cartilage (Figure 5). The

sequence used to obtain the image reported in Fig. 5

was a 3D FIESTA with 0.156 mm in-plane

resolution, FA = 20, TR = 6.3 ms, TE = 2.5 ms,

thickness = 0.8 mm.

The patellar cartilage was then segmented by

means of the ITK-SNAP (Yushkevich, 2006)

software tool, obtaining the results reported in

Figure 6 (in the three axial planes and in the 3D

volume rendering). Then the volume of the

segmented cartilage has been quantified, obtaining

the result of 1779 mm

Figure 5: Knee of a healthy volunteer in which clearly

depicts the trabecular architecture of the bone and the

cartilages.

Figure 6: Segmentation and 3D rendering of the patellar

cartilage of a healthy volunteer.

The knee RF coil was also used to obtain images

of the knee in the pathological volunteer, in which

the pathological change is clearly evident.

The sequence used to obtain the image reported

in Figure 7 (left knee) was a 3D FIESTA with 0.156

BIODEVICES 2016 - 9th International Conference on Biomedical Electronics and Devices

232

mm in-plane resolution, FA = 20, TR = 6.3 ms, TE =

2.4 ms, thickness = 0.8 mm.

The sequence used to obtain the image reported

in Figure 8 (right knee) was a 3D FIESTA with

0.156 mm in-plane resolution, FA = 20, TR = 9.0

ms, TE = 3.4 ms, thickness = 0.8 mm. In Figure 9 is

reported an example of T2* map calculation of the

same slice of Figure 8, realized by means of a 3D

MERGE sequence with 6 echo times. In these

images, of the right knee, the lack of cartilage in the

medial facet of the patella and other alterations of

the cartilage are clearly highlighted.

Figure 7: Left knee of the pathological volunteer.

Figure 8: Right knee of the pathological volunteer.

Figure 9: T2* map of the right knee of the pathological

volunteer.

4 CONCLUSIONS

We presented the recent achievements in human

MRI with a 7 T MR whole-body scanner. We

designed and developed dedicated RF surface coils

for

H imaging. The RF coil prototype for neck has

been validated in vivo on healthy volunteers. The

obtained results in the cervical spinal cord

demonstrate a high image quality and if confirmed

in a larger sample of patients might constitute a

promising tool in exploring a complex anatomical

region that is prone to susceptibility artifact at UHF.

The RF coil prototype for knee has been validated in

vivo on healthy and pathological volunteers. Due to

the low amount of water in the bone and the small

size of the cartilage, this district is well suitable for a

study using UHF. With this technology it was

possible to carry out a thorough assessment both

morphological and functional. The obtained results

demonstrate that the research in trabecular bone and

cartilages characterization, comprising quantitative

assessment of cartilage volume and evaluation of

biochemical behaviour, can take advantage from

UHF MR with dedicated coils.

REFERENCES

Farooki, S, C.J. Ashman, CJ et al., 2002. “In vivo high-

resolution MR imaging of the carpal tunnel at 8.0

Tesla,” Skeletal Radiology, vol. 31(8), pp. 445-450.

Kraff, O, Fischer, A et al., 2015. “MRI at 7 Tesla and

Above: Demonstrated and Potential Capabilities,”

Journal of Magnetic Resonance Imaging, vol. 41, pp.

13–33.

Krug, R et al., 2009. “Imaging of the Musculoskeletal

System In Vivo Using Ultra-high Field Magnetic

Resonance at 7 T”, Investigative Radiology, vol. 44 n.

9, pp. 613-618.

Gambarota, G, Veltien, A, et al., 2007. “Magnetic

resonance imaging and T2 relaxometry of human

median nerve at 7 Tesla,” Muscle Nerve, vol. 36(3),

pp. 368-373.

Majumdar, S, 2008. “Ultra-High Field-7 Tesla Imaging of

the Musculoskeletal System,” ISMRM High Field

Workshop, Rome.

Mispelter, J, Lupu M et al., 2006. “NMR Probeheads for

Biophysical and Biomedical Experiments: Theorical

Principles Pratical Guidilines,” Imperial College Press.

Retico A, Stara, R et al., 2015. “Non-invasive assessment

of Neuromuscular Disorders by 7 tesla Magnetic

Resonance Imaging and Spectroscopy: Dedicated

radio-frequency coil development”, 2015 IEEE

International Symposium on Medical Measurements

and Applications (MeMeA) pp. 68-73.

Magnetic Resonance Imaging at 7 Tesla with Dedicated Radiofrequency Coils - Application to Cervical Cord and Knee

233

Regatte, RR and Schweitzer, ME, 2007. “Ultra-high field

MRI of the muscoloskeletal system at 7.0 T,” Journal

of Magnetic Resonance Imaging, vol. 25(2), pp. 262-

269.

Smith, P, 1995. Electronic Applications of the Smith

Chart, 2nd edition, 1995, SciTech/Noble Publishing,

Raleigh, North Carolina.

Ugurbil, K, Adriany, G et al., 2003. P. Andersen, et al.,

“Ultrahigh field magnetic resonance imaging and

spectroscopy,” Magn. Reson. Imaging, vol. 21, pp.

1263-1281, 2003.

Yushkevich, PA, Piven, J et al., 2006. “User-guided 3D

active contour segmentation of anatomical structures:

Significantly improved efficiency and reliability.”

Neuroimage vol. 31, n. 3, pp 1116-28.

BIODEVICES 2016 - 9th International Conference on Biomedical Electronics and Devices

234