Simultaneous EEG-fMRI at 7T: Radiofrequency (RF) Heating and Shielding of a novel EEG cap

2340 

Abstract Submission 

Rebecca Meagher¹, David Carmichael², Özlem Ipek³, Tracy Warbrick⁴

¹King's College London, London, United Kingdom , ²King's College London, London, Please select an option below, ³King's College London, London, London, ⁴Brain Products GmbH, Gilching, Gilching

Rebecca Meagher  
King's College London
London, United Kingdom

David Carmichael  
King's College London
London, Please select an option below

Özlem Ipek, Doctor  
King's College London
London, London

Tracy Warbrick, Doctor  
Brain Products GmbH
Gilching, Gilching

The simultaneous acquisition of EEG and fMRI (EEG-fMRI) at 7T can provide very high temporal (sub-millisecond) resolution and spatial (sub-millimetre) resolution allowing non-invasive investigation of human brain function at the meso-scale. However, performing EEG-fMRI at ultra-high field comes with significant challenges because shorter RF wavelengths can facilitate increased coupling to the EEG cap (Phong Lê, Gruetter, Jorge, & Ipek, 2022), while also increasing the Specific Absorption Rate (SAR) produced by electric fields. In this study, we investigated a novel EEG cap incorporating segmented resistors, designed to avoid RF shielding at 7T, by measuring heating and transmit field (B₁⁺) changes.

Data was acquired in 3 healthy volunteers. EEG measurements were recorded using the 32-channel 7T prototype cap with 4 carbon wire loops (BrainCap-MR7 FLEX, Brain Products GmbH). This novel cap design includes 5kOhm resistors placed on the flexprint PCB with another set of resistors on the EEG amplifier connection point to mitigate the formation of RF standing waves on long EEG cabling. (Phong Lê, Gruetter, Jorge, & Ipek, 2022) In addition, this design allows the modular connection between the EEG cap and the amplifiers through the 7T single-Tx birdcage coil (Nova Medical Inc., MA, USA) where the MR data was acquired at 7T (MAGNETOM Terra, Siemens Healthcare). Temperature probe measurements were acquired throughout the duration of the scan (Omniflex2, Neoptix, CA), with a sensitivity of +/- 0.1 ºC, for each volunteer with 2 probes being placed underneath different EEG cap electrodes, and 1 placed on the ExG amplifier upper surface. The temperature data was sampled at a rate of 1Hz , and this data was subsequently smoothed with a 5s box kernel. For volunteers 1 and 2, the most lateral electrodes (FT9 and FT10) with the longest cabling, and therefore, the most susceptible to heating, were selected for temperature sampling. For volunteer 3, probes were placed at electrodes at the posterior region of the cap (TP9 and O1) as these are the electrodes closest to the coil Thermal images were captured (Seek Shot Thermal Camera, SeekThermal, CA, USA) before and after the scan. The specific pulse sequences used (localiser, B₁⁺ maps, B₀ maps 2D-TSE and MP2RAGE) and their SAR are described in Figure 1(b). At the end of the protocol the cap was removed and the subject repositioned within the coil then the localiser, B₁⁺ maps, B₀ maps were repeated.

Figure 1(a) shows an overview of the temperature results. For all 3 volunteers, the temperature is seen to remain stable at both electrodes and at the ExG amplifier as both a function of time throughout the duration of the scan. There is minimal heating associated with scan onset apparent even for sequences with maximum normal mode SAR (TSE and MP2RAGE). Slow temperature drifts that are not related to scanning are visible, likely caused by natural variations in the thermal environment of the room, the cap, and the subjects head. These findings are reinforced by the thermal camera images, whereby there is no discernible difference between images captured before and after the scanning session, and no clear locations of increased heating. Figure 2 displays the B₁⁺ maps for both the EEG cap on and cap off condition. Across all 3 volunteers, similar B₁⁺ maps were obtained for both the cap on and cap off condition, with minimal visible RF reduction observed.

No measurable heating at the EEG electrodes or ExG amplifier was detected in any of the volunteers in this study, providing evidence that the risk of significant heating is low. Moreover, no visible RF reduction was observed between the EEG cap on and cap off conditions.

BOLD fMRI ¹

EEG ²

Multi-Modal Imaging

Electroencephaolography (EEG)

FUNCTIONAL MRI

HIGH FIELD MR