AUTOMATIC DEACTIVATION DESIGN FOR PHASED ARRAY

SURFACE PROBE IN 1.5T MRI

Fotios N. Vlachos, Anastasios D. Garetsos and Nikolaos K. Uzunoglu

School of Electrical and Computer Engineering, Narional Technical University of Athens

9 Iroon Polytechniou, 15773, Zografou

Keywords: Automatic tuning, phased array, MR imaging, spectroscopy.

Abstract: We have designed and developed an automatic switching mechanism that deactivates and activates a

reception coil during the MR experiment according to the phase it is at. The mechanism uses a feedback

loop in which a comparator defines whether the current reception signal derives from the RF excitation

pulses or the MR signal and then triggers an analog switch at the back-end of the coil accordingly. We

applied the mechanism on a custom-made four channel phased array probe and tested its functionality by

transmitting RF pulses to the probe of similar length and power to those used in actual MRI systems. The

results presented in this paper demonstrate the robustness of the design and its switching accuracy.

1 INTRODUCTION

In the last ten years there has been much progress in

the development of fully autonomic probes for MR

Imaging and Spectroscopy. In most of the

experimental attempts, emphasis is given on the

automatic tuning and matching (Hwang and Hoult,

1998; Pérez de Alejo et. al., 2004) of the coils in

order to improve the signal-to-noise ratio (SNR)

values and accelerate the initialization procedures

that keep the patient for an extensive period of time

in the MRI bore.

All automatically tuned and matched coils

require being compatible with the pulse sequences

used in the MR experiment, which implies detuning

of the coils during the RF pulse transmission and re-

tuning for the MR signal reception. A series of

complex automatic deactivation techniques have

been developed and tested in the past (Venook et.

al., 2005), which function in parallel with the tuning

and detuning procedures but suffer robustness and

poor results.

The most common deactivation technique that

has been applied in both conventional and

experimental non-automatic configurations is the use

of PIN diodes at the back-end of the probe (Yung et.

al., 2003; Barberi et. al. 2000). These configurations,

however, are totally dependable on the external

signals that the MR scanner supplies in order to turn

on or off the PIN diode.

In this study, we present a simple and robust

automatic design that undertakes the responsibility

of deactivating and activating the probe during the

RF pulse transmission and the MR signal reception

phase respectively. The design does not require the

presence of any external signals and is fully

functionable with a large variety of RF pulse lengths

and powers.

2 MATERIALS AND METHODS

The automatic deactivation circuitry that was

constructed was applied on a prototype human

prostate phased array probe, which we designed and

developed after simulation modelling and

laboratorial measurements.

2.1 Probe Design

The probe is consisted of four rectangular coils of

dimensions 8 × 16 cm and the material that was used

for their construction was copper tape 1 cm wide and

1 mm thick. The coils are distributed into two pairs

of adjacent elements and each pair is positioned

inside an orthogonal shaped conductor frame made

of acetal (Fig. 1A). When a patient is examined the

two frames are locked at a fixed distance so the

elements are placed at the center posterior and

160

N. Vlachos F., D. Garetsos A. and K. Uzunoglu N. (2008).

AUTOMATIC DEACTIVATION DESIGN FOR PHASED ARRAY SURFACE PROBE IN 1.5T MRI.

In Proceedings of the First International Conference on Biomedical Electronics and Devices, pages 160-163

DOI: 10.5220/0001050501600163

 SciTePress

anterior surface of the pelvic area in order to achieve

optimum phased array performance.

Magnetic field and inductance calculations were

carried out using the Biot-Savart integral expression

(Wright and Wald, 1997). The configuration of the

probe was modelled with simulation programming

in order to measure the magnetic field’s intensity in

all three dimensions (Fig. 2A) and calculate the

intensity’s drop percentage at the center of the pelvic

region in comparison to the intensity 1 cm away

from the coils at the surface of the test object. The

theoretical measurements were done for various

distances between the frames of the probe and the

Figure 2: Probe’s theoretical simulation results. (A) Computed magnetic field’s intensity distribution on various xy planes

for frames distance 24 cm. (B) Magnetic field’s intensity at the line that connects the centers of the parallel frames fo

various frames distances.

MRI

Coaxial

Cable λ

4:16 Balun

Coil 1

Coil 2

AD8307

Frames

Distance

Coil 1 Coil 2

Coil 3 Coil 4

ZC832

:RF

+5V

RSW2-25P

+5V

COMPAR_OUT

+30V

PIC16F877

(ADC)

Log Detector

Analog Switch (TTL)

ZC832

ADC_VREF

Threshold

+12V +12V

Figure 1: Phased Array probe’s design and circuit diagram. Two pairs of coils are positioned inside two acetal frames (A).

Each coil carries two antiparallel crossed diodes for passive blocking and a variable capacitor for tuning (B). The fine-

tuning/matching circuitries are positioned on four PCBs at λ distance away from the elements. Each PCB includes the

automatic deactivation circuitry that interacts with the probe through a feedback loop. The tuning section (C) uses two pairs

of ZC832 varactor diodes in a pi-network to match the output impedance. The feedback section (D) rectifies the RF pulses

into DC signals, which are then compared to a threshold DC value. The comparator’s output is processed in the

microcontroller (E), which determines when the probe is in the activation and the deactivation phase and controls an analog

switch that connects the probe to the MRI scanner.

AUTOMATIC DEACTIVATION DESIGN FOR PHASED ARRAY SURFACE PROBE IN 1.5T MRI

161

results showed that the drop percentage does not fall

under 7.66% in the worst case scenario of frames

distance 30 cm (Fig. 2B).

The method that was used for the adjacent

elements decoupling was overlapping (Roemer et.

al., 1990). Theoretical calculations with simulation

programming indicated that the distance the adjacent

coils should have in order to minimize the mutual

inductance is 6.7 cm. That translates in 8-6.7=1.3 cm

coils overlap.

The probe is enhanced with a passive blocking

network of two anti-parallel high speed diodes

(Fig. 1B) that behave as a short circuit when they are

forward-biased and serve as a safety precaution for

the rest of the system’s electronics (Noeske et. al.

2000; Zhang and Webb, 2005). 4:16 balun elements

at the back-end of the rectangular loops convert the

balanced output signal of the coils to 50 Ohm

unbalanced. The tuning and the matching of the

probe is manually controlled from a variable non-

magnetic capacitor and 2 pairs of varactors in a pi-

network (Fig. 1C), which lies one wavelength (λ)

away from the coils.

2.2 Automatic Deactivation Design

The function of the MR probe is divided into two

phases: the RF pulse transmission phase, during

which the probe should be deactivated and

disconnected from the MR scanner and the MR

signal reception phase, during which the MR probe

should be activated. The transition of the probe from

the deactivation to the activation phase is controlled

by the automatic deactivation circuitry, which

connects between the probe and the scanner’s

preamplifier. The design of the circuitry is based on

a feedback loop, which uses the probe’s reception

signals to define the phase that it is at.

An analog switch (RSW2-25P) is used to block

the output of the probe from connecting to the MR

scanner, when the RF pulses are transmitted. The

switch is triggered by two signals (C1, C2) that a

microprocessor sends (Fig. 1E). When the C1 signal

is on (activation phase), the RF1 position of the

switch is short-circuited and the scanner receives the

MR signal. Contrarily, when the C2 signal is on

(deactivation phase), the RF2 position is short-

circuited and the high-power transmitted RF pulses

are grounded, protecting the scanner’s preamplifier.

The decision between triggering signal C1 or C2 is

taken from a comparator (LM393AD), which

compares a pre-defined threshold DC voltage with

the output DC signal that is rectified from the

transmitting RF pulses, using a log detector

(AD8307) and two RC low-pass filters (Fig. 1D).

Consequently, when the high power RF pulses are

transmitted, then the output DC signal’s amplitude is

higher than the threshold voltage, the analog switch

is turned off and the probe enters the deactivation

phase (C2 signal triggering). Contrariwise, when the

output DC signal’s amplitude is lower than the

amplitude of the threshold, the analog switch is

turned on and the probe is re-activated (C1 signal

triggering), receiving low power MR signal from the

hydrogen molecules’ resonance.

10μs

OFFOFF

Outpu

ON ON ON

DEACTDEACT ACT ACT ACT

Analog Switch

10μs

Phase

RF Rectification

DC Threshol

Inpu

Figure 3: Switching mechanism’s experimental results. (A) Transmitted RF pulses that are used as input to define the

experiment’s phase. (B) DC pulses produced from log detector’s RF rectification (lower line in first combiscope figure).

the phase. (D) Optimum analog switch’s behaviour following the experiment’s phases. (E) Actual analog switch’s

behaviour with the undesired latency. Upper line in both combiscope figures depicts the trigger that generates the RF pulse.

BIODEVICES 2008 - International Conference on Biomedical Electronics and Devices

162

3 RESULTS

Before the testing of the automatic deactivation

circuitry we tuned and matched the coils of the

probe in the Larmor frequency (63.87 MHz) by

applying an average human pelvic region load on the

frames and adjusting the values of tuning

components. Using the network analyzer

(HP8719D) to measure the reflection coefficient we

managed to drop the S

parameter at -55 dB,

keeping the resonance frequency range below 250

KHz, which led to very accurate tuning. The

decoupling between the adjacent elements was also

successful, since the transmission coefficient S

drops below -30 dB.

The functionality of the analog switch was tested

in the laboratory using a Signal Generator (HP ESG-

4000A) and a Combiscope (FLUKE PM3380B). RF

pulses of the same power and length with those

transmitted from the MRI system were created in the

Signal Generator and were sent to the probe as input.

The RF pulses varied in length from 2-5 ms and in

power from 5-20 dBm.

The first set of measurements examined the log

detector’s functionality. Specifically, we measured

the DC signal produced from the RF pulse

rectification (Fig. 3B). The resulting DC pulse is

initiated and terminated almost immediately after the

beginning and the end of the signal generator’s RF

trigger respectively. Also, the correspondent DC

pulse’s amplitude is equal to the RF pulse’s

amplitude as expected, allowing accurate

comparison with the DC threshold.

The second set of measurements showed the

output of the probe and verified the turning off and

on of the analog switch during the activation and the

deactivation phase respectively (Fig. 3C). A

potential disadvantage of the method is that there is

an undesired latency of 10 μs in the switching

process that is capable of producing artifacts in the

imaging data (Fig. 3E). The latency is caused mainly

from the processing delays of the microcontroller

that triggers the analog switch and remains constant

without regard to the RF pulse length and power that

is triggered.

4 DISCUSSION

Certain improvements could be applied on the

automatic switching mechanism of the circuitry in

order to overcome the presence of latency in the

function of the analog switch. A way to reduce the

latency is to control the switch directly from the DC

signal that derives from the comparator’s output,

bypassing the time-consuming processing of the

microcontroller.

Also, a practical problem could potentially

appear in the clinical application of the automatic

deactivation circuitry. The probe detunes itself

automatically during the RF pulse transmission and

does not require a decoupling signal from the

scanner. However, many MRI scanners’ protocols

run primary tests on the connected probes by

sending pulse signals in the opposite direction for

software initialization. In that case, the switch would

cause compatibility issues and the probe would not

be recognized by the MR system.

Our prototype automatic deactivation design is a

robust and simplified mechanism that can be applied

on self-tunable MR coils. It was tested in various

conditions and found to be fully functional and able

to switch off or on the probe at all times.

REFERENCES

Hwang, F., Hoult D.I. 1998, Automatic Probe Tuning and

Matching, Magnetic Resonance in Medicine, 39,

pp. 214-222.

Pérez de Alejo, R., Garrido, C., Villa, P., Rodriguez, I.,

Vaquero, J.J., Ruiz-Cabello, J., Cortijo, M., 2004,

Automatic Tuning and Matching of a Small

Multifrequency Saddle Coil at 4.7T, Magnetic

Resonance in Medicine, 51, pp. 869-873.

Venook, R.D., Hargreaves, B.A., Gold, G.E., Conolly,

S.M., Scott, G.C., 2005. Automatic Tuning of Flexible

Interventional RF Receiver Coils, Magnetic

Resonance in Medicine, 54, pp. 983-993.

Yung, A.C., Oner, A.Y., Serfaty, J.M., Feneley, M., Yang,

X., Atalar, E., 2003. Phased-Array MRI of Canine

Prostate Using Endorectal and Endourethral Coils,

Magnetic Resonance in Medicine, 49, pp. 710-715.

Barberi E.A., Gati J.S., Rutt B.K., Menon R.S., 2000. A

Transmit-Only/Receive-Only (TORO) RF System for

High-Field MRI/MRS Applications, Magnetic

Resonance in Medicine, 43, pp. 284-289.

Wright S.M., Wald L.L., 1997. Theory and Application of

Array Coils in MR Spectroscopy, NMR in

Biomedicine, 10, pp. 394-410.

Noeske R., Seifert F., Rhein K-H., Rinneberg H., 2000.

Human Cardiac Imaging at 3 T Using Phased Array

Coils, Magnetic Resonance in Medicine, 44, pp. 978-

982.

Zhang, X., Webb, A., 2005. Design of a Four-Coil Surface

Array for in Vivo Magnetic Resonance Microscopy at

600 MHz, Concepts in Magnetic Resonance Part B,

24B(1), pp. 6-14.

Roemer P.B., Edelstein W.A., Hayes C.E., Souza S.P.,

Mueller O.M., 1990. The NMR Phased Array,

Magnetic Resonance in Medicine, 16, pp. 192-225.

AUTOMATIC DEACTIVATION DESIGN FOR PHASED ARRAY SURFACE PROBE IN 1.5T MRI

163