Back to Projects

AI-enabled Imaging, Emerging Imaging

Investigating the epileptic brain with simultaneous EEG and MRI at 7 Tesla

Project ID: 2021_017

1st supervisor: Ozlem Ipek, King’s College London
2nd supervisor: David Carmichael, King’s College London
Clinical Champion: Alexander Hammers, King’s College London

Aim of the PhD Project:

  • Designing novel solutions into EEG design and MR hardware to improve simultaneous EEG/fMRI data.
  • Develop an AI based algorithm to perform high-quality motion-free data collection with the advantage of (following training) dramatically increasing its computational speed.
  • Demonstrate feasibility for clinical studies in epilepsy patients.

Project Description / Background:

Epileptic seizures originate from structural and electrical pattern changes in the brain, but the mechanism related to seizure networks is not well known. Simultaneous scalp electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) acquisition allows the measurement of brain activity with large-scale network organization at high spatial-temporal resolution. MRI at 7T enables a large increase in sensitivity to fMRI signal changes. This sensitivity can be utilised to either to measure haemodynamic responses to epileptic activity much more accurately either in the spatial domain using MRI and temporal domain using EEG [1].

However, there are a number of engineering challenges that need to be overcome to combine MRI with EEG. Firstly, electrode heating becomes more likely owing to both the reduced wavelength of the RF compared to wire lengths [2] and the increased RF power required to achieve the same pulse sequence at 7T compared to 3T [3]. Work is required to develop a new EEG configuration that is compatible with the existing 7T RF coils, minimises EEG artefacts and maximises safety by reducing induced voltages. This will be achieved by altering wiring configurations and wire impedance for example by distributing impedance to minimise effective wire lengths.

Secondly, EEG quality is degraded by an increased sensitivity to both pulsatile and bulk head motion. We will build on a recently awarded project grant (GOSHCC) that will develop individualised head cushions for paediatric 7T MRI to minimise motion by including EEG wiring and electrodes into the customised cushion design. The safety of these new configurations requires a combination of simulation and experimental evaluation with existing MR-compatible 64-channel scalp EEG systems [4]. In addition, retrospective motion correction methods will be applied for correcting small motion in 2D and 3D MRI sequences with signal-to-noise and k-space ordering. Finally, the methods will be investigated to acquire high quality EEG and MRI data using implemented AI algorithms.

Thirdly, a clinical diagnostic protocol will be developed to acquire functional network data which allows characterization of simultaneous brain activity of linked areas in the brain in patients with epilepsy. As a result of employing functional MRI, diffusion MRI and spectroscopy, an optimal simultaneous MRI and EEG quality will improve characterization of the mechanism of epileptic seizure networks.

There is a general research ethics approval submission in place for the 7 Tesla at St Thomas’ Hospital which will be extended to including any hardware built in-house.  We will follow the 7T MRI safety SOP (both supervisors are involved in the local committee) before using the EEG device on humans. Before using it on humans, its function and safety will be validated by computer simulations and simultaneous MRI-EEG measurements on MRI test objects with temperature probe sensors. Therefore, most of the technical development can be achieved without human participants. The supervisors have extensive experience of testing and imaging using in-house hardware such as RF coils and imaging patients with EEG systems [4,5]. We include a placement at Nottingham’s Sir Peter Mansfield Imaging Centre where EEG-fMRI has been performed at 7T.


[1] Centeno M., Tierney T.M., Perani S., Shamshiri E.A., StPier K., Wilkinson C., Konn D., Vulliemoz S., Grouiller F., Lemieux L., Pressler R.M. Combined EEG-fMRI and ESI improves localisation of paediatric focal epilepsy. Ann. Neurol. 2017; 82 (2), 278–287.
[2] Le T.P., Ipek Ö., Jorge J., Gruetter R. Improved transmit field homogeneity in simultaneous EEG-fMRI at 7T by increasing EEG wire resistance. Proc. Intl. Mag. Reson. Med. 2018; 26, 4314.
[3] Hawsawi H.B., Carmichael D.W., Lemieux L. Safety of simultaneous scalp or intracranial EEG during MRI: A review. Frontiers in Physics, 2017; 5(42).
[4] Jorge J., Grouiller F., Ipek Ö., Stoermer R., Michel C.M., Figueiredo P., van der Zwaag W., Gruetter R. Simultaneous EEG-fMRI at ultra-high field: Artifact prevention and safety assessment. Neuroimage 2015; 105, 132-144.
[5] Carmichael D.W., Thornton J.S., Rodionov R., Thornton R., McEvoy A., Allen P.J., Lemieux L. Safety of localizing epilepsy monitoring intracranial electroencephalography electrodes suing MRI: radiofrequency-induced heating. J. Magn Reson Imaging 2008; 28(5), 1233-1244.

Figure 1: Alpha power variation when an eyes-open/eyes-closed task captured during simultaneous EEG-fMRI (top image and right image acquired by EEG and brain images below acquired by 7T fMRI).

Back to Projects