A Patient-adjustable MRI coil for implant-friendly imaging of deep brain stimulation: Design, construction, and patient-specific numerical simulations
Laleh Golestanirad1, Boris Keil1, Maria Ida Iacono2, Giorgio Bonmassar1, Leonardo M Angelone2, Cristen LaPierre1, and Lawrence L Wald1

1Radiology, Massachusetts General Hospital, Charlestown, MA, United States, 2Division of Biomedical Physics, Office of Science and Engineering Laboratories, Center for Devices and Radiological Health, US Food and Drug Administration, Silver Spring, MD, United States

Synopsis

Recently we presented the feasibility study of using a reconfigurable DBS-friendly head coil, composed of a patient-adjustable rotating birdcage transmitter, and an integrated 32-channel receiver array to reduce SAR during imaging of patients with deep brain stimulation implants. Here we introduce the first prototype of such coil system, and present results of finite element simulations on patient-derived numerical models of realistic DBS lead trajectories, which characterize its SAR reduction performance.

Introduction

Each year approximately 300,000 patients with medical implants including deep brain stimulation (DBS) devices are denied magnetic resonance imaging (MRI) examination due to safety concerns 1. Recently we presented results of a feasibility study for using reconfigurable DBS-friendly transmit/receive head coils, composed of a patient-adjustable rotating birdcage transmitter, and an integrated 32-channel receiver array at 1.5T 2. Here we present the first prototype of such coil system, along with the results of measurements and comprehensive numerical simulations that characterize its image quality and specific absorption rate (SAR) profile.

Materials and Methods

Hardware development and safety tests

The coil system was composed of a linearly-polarized rotating birdcage transmitter, and an anthropomorphic 32-channel receive array at 1.5 T (see Fig.1). The transmitter has a slab-like region of low electric field that is systematically adjusted for each patient to encompass the DBS implant. This technique reduces the local SAR at the implant tip by 500 times below that of a circularly polarized (CP) birdcage. An array of 32-channel surface receiver coils was designed for maximum reception. Details of array design and construction can be found in 3. The ensemble of transmit-receive array went through a comprehensive battery of safety tests to assess the quality of active detuning and SNR maps, as well as evaluating temperature increases (both in the coil’s circuitry due to eddy-currents, and in a head phantom due to RF absorption) during long continuous scans (~15 min).

Numerical models

The coil’s SAR reduction performance was assessed for both generic DBS lead trajectories and patient-derived realistic lead models. Generic DBS leads models were integrated into a 1×1×1mm3 numerical head model with 15 different tissue classes. Details of segmentation and construction of the head model are given in 4. The location of the sub-thalamic nucleus (STN), —one of anatomical structures targeted for DBS, was estimated from the MRI data upon which the head model was built. After the STN was located, a series of entry points were selected on the skull based on typical angles of approach for the STN DBS (~60˚ in the anterior-posterior direction and 15˚–30˚ from the sagittal plane 5). The rest of the lead was placed in the subcutaneous structure along anterior-posterior position. To evaluate the SAR reduction performance in practical cases, leads with the extra-cranial portion looped around the surgical burr hole were also modeled.

Patient-specific simulations were performed on four realistic DBS lead trajectories, extracted from post-operative CT images of patients operated on at Massachusetts General Hospital. The lead trajectories were manually segmented from CT images and a polyline representing lead trajectory was constructed and exported to the electromagnetic solver (HFSS, ANSYS Inc). Next, a DBS lead model including 4 cylindrical contacts (“electrodes”) connected through a solid core surrounded by hollow insulation was constructed in HFSS. Finally, homogeneous head models were built from the CT data, and were assigned the electric properties of grey matter at 64 MHz.

For both generic and realistic cases, FEM simulations were performed for the full range of coil rotation angles and 1g-averaged SAR was calculated using HFSS built-in SAR module.

Results and Future Work

The ensemble of transmit/receive coil showed less than 3˚C temperature rise in coil components during test scans. Images showed increased SNR up to 5 fold in cortical structures, and up to 2 fold at the level of deep brain nuclei with respect to a circularly polarized (CP) birdcage (see Fig. 1). In all simulations including (1) generic DBS lead trajectories with wire segments that were 15˚–30˚ out of plane (Fig. 2c), (2) leads with looped segments (Fig. 2d), and (3) patient-derived realistic lead trajectories (Figs. 3 and 4), an optimum coil angle was found that reduced the maximum local SAR at electrode tip below the maximum SAR value in the head. The location of this optimum position was however dependent on the individual lead trajectory, indicating the need to have a patient-adjustable transmitter.

In the last step, the sensitivity of the results to electrical conductivity of the anatomical structure surrounding the DBS (labeled as “grey matter” in the numerical model) was evaluated. Simulations were performed with a range of conductivities from 0.01-3 S/m, and with different coil rotation angles. We found that when the coil was in the optimum position, the maximum SAR values were not sensitive to electrical conductivity of the structure surrounding the electrode (see Fig. 5).

We are now in the process of evaluating a cohort of 20 DBS patients to perform uncertainly budget calculations assessing the safety of this new technology for post-operative imaging of DBS patients.

Acknowledgements

This work was partially supported by NIH grant P41EB015896.

References

1. Shellock FG, Spinazzi A. MRI safety update 2008: part 2, screening patients for MRI. American Journal of Roentgenology 2008;191(4):1140-1149.

2. Golestnirad L, Keil B, Bonmassar G, Mareyam A, Wald LL. A rotating transmit coil and 32ch receive array for high-resolution brain imaging of DBS patients. 2015; Toronto, Canada.

3. Keil B, Wald LL. Massively parallel MRI detector arrays. Journal of Magnetic Resonance 2013;229:75-89.

4. Makris N, Angelone L, Tulloch S, Sorg S, Kaiser J, Kennedy D, Bonmassar G. MRI-based anatomical model of the human head for specific absorption rate mapping. Medical & biological engineering & computing 2008;46(12):1239-1251.

5. Malone DA, Dougherty DD, Rezai AR, Carpenter LL, Friehs GM, Eskandar EN, Rauch SL, Rasmussen SA, Machado AG, Kubu CS. Deep brain stimulation of the ventral capsule/ventral striatum for treatment-resistant depression. Biological psychiatry 2009;65(4):267-275.

Figures

Figure 1: Top row: The ensembles of reconfigurable transmitter and 32-ch receive array with an anthropomorphic head phantom. Bottom row: Noise covariance matrix and SNR maps

Figure 2: (a) Heterogeneous head model with 15 tissue classes (b) Location of STN was estimated in the head model and a series of entry points were selected on the skull with different approach angles. (c)&(d) 1g-averaged SAR profiles for different coil rotation angles.

Figure 3: patient-derived homogeneous head models and DBS implants.

Figure 4: Maximum 1g-averaged SAR for patient-derived models of Figure 3. For all models, an optimum coil angle was found that reduced the SAR at least 500 times.

Figure 5: Maximum SAR and background SAR for 1g averaged SAR inside the heterogeneous head model for a range of gray matter electrical conductivities. Left: coil in default position (feed up), Right: coil with optimum rotation angle for SAR reduction.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0354