Rung Sensors for Birdcage Coils


The goal of this project is to study the current pattern on an MRI transmission coil (antenna) in real time, with the aim of detecting anomalous patterns that may be indicative of coil loading conditions creating a safety hazard to the patient. Such issues could lead to the burning of patients, and are due to high capacitive coupling to tissue, or resonant coupling to implanted devices such as pacemaker leads or neurostimulators.


The basic measurement of the rung current is achieved through inductive coupling of the rung conductor to a pickup loop, in the form of a slotted planar coil transformer. The signal from the pickup loop is then used to modulate the bias of a photodiode, which transmits the signal on an optical fiber.


The receiver unit for this project has the task of multiplexing the 8 signals from the different coil rungs. The signals are multiplexed using 3 4:1 CMOS multiplexers, and are then passed through two amplification stages - a transimpedance amplifier and a variable gain amplifier. Triggers arrive from the MRI console, and a microcontroller on board (ATMega328p) uses this to time-synchronize the acquisitions, and manage the multiplexers and VGA.

Conference Abstract

Below is an abstract I have submitted to the ISMRM conference for 2014.

Sensing of Birdcage Rung Currents for Detection of Anomalous Loading

Introduction A primary source of risk to the MRI subject is interaction with the radiofrequency fields. Typically transmitted by a birdcage coil, well-known hazards include the coupling of the coil to implanted or interventional wires, and capacitive coupling to tissue of a patient touching the bore surface. Both have been known to cause severe burns. It has also been shown [1,2] that by monitoring the impedance spectrum of the birdcage coil at its drive port, characteristic features appear which are indicative of resonant coupling. Here we propose a new measurement, which can be performed quickly and at low power, wherein we measure the current flowing on each rung of the birdcage coil. Information contained in the spatial current distribution would not necessarily be available at the drive port, and may be indicative of anomalous loading conditions.
Methods. In order to dynamically determine the current flowing in each rung of the birdcage coil, the individual rungs have been instrumented with optically coupled current sensors (Figure 1). The sensors are based on the slotted planar coil current sensors described in [1], which implement a 1:1 transformer in a 4-layer PCB. The output of the transformer is capacitively coupled to a biased photodiode (Firecomms FC300T), outputting the RF-modulated signal over plastic optical fiber (POF). The circuit is powered by a non-magnetic LiPo battery, and includes a battery-charging IC. The receive circuit includes 8 optical receivers (Figure 2) and RF multiplexers which select and output the signal to a transimpedance amplifier, which is followed by a variable gain amplifier and a Medusa console. The multiplexing and gain control is managed by a microcontroller, and a different rung sensor is interrogated on each TR. Current and image data are obtained concurrently. To calibrate the sensors, a hose phantom containing a saline solution is wrapped around the outer diameter of the coil to facilitate B1 mapping in the vicinity of the rungs. A double-angle field map is created (Figure 3, TR = 1500ms, FOV = 50.8cm) and a contour is drawn around the rung, in a region where the flip angle is close to 45 degrees. Since there is only signal on one side of the rung, the full integral is estimated assuming a symmetric profile. To test the effects of loading on the pattern of rung currents, a rotatable phantom was constructed, comprising a 1.9 L bottle of saline solution and optionally a wire phantom, which is a wire in a tubular structure, also filled with saline solution.
Results Figure 3 shows the result of the sensor calibration. The dashed line shows the currents estimated from the image, and the solid lines indicate the magnitude (length) and phase (direction) of the sensor measurements. Even without calibration, the sensor readings show a rough correspondence in magnitude to image-estimated currents. Currents in rungs 1 and 8 were too small to be measured accurately by the imaging method and calibration coefficients were estimated by averaging coefficients of other rungs. As expected, the phase of the current on opposite sides of the coil is opposite. Figure 4 shows the currents measured on the rungs of a coil loaded by the bottle and wire phantom. The dominant effect is a uniform reduction in rung currents, to be expected from coil loading. A distortion in the current pattern is observed, however, when the wire phantom is close to rungs carrying a high current.

Conclusion. Observed deviations in the current pattern were found to be small but distinct. Further work could be done with a broader range of loads to identify characteristic distortions, including in higher field systems where effects should be more pronounced. In addition, these measurements could be used as an aid in B1 estimaton and shimming schemes.

References. [1] Stang et al., Proc. 18th ISMRM, p44, 2010 [2] Ellenor et al., ISMRM Safety Workshop, 2012. NIH Grant Support: R01EB008108, R33CA118276, R21EB007715, P01CA159992.