The purpose of this project is to develop the method of combined recording of electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) in order to improve understanding of the relationship between electrical (EEG) and blood flow (fMRI) responses to epileptic discharges as a necessary step towards clinical use. One factor that currently limits the use of EEG-fMRI in patients with epilepsy is that a relatively large proportion of patients do not show any fMRI response despite epileptic activity having been detected on the EEG recorded during scanning. The reasons for this are unclear, which makes it difficult to predict in advance whether useful information will be gained from the scanning session. What is it about some epileptic discharges that results in a detectable change in the fMRI signal, while others which are no less obvious or frequent do not? This question will be addressed by determining the factors that are responsible for the occurrence of robust fMRI signal changes via detailed morphological, topographical and spectral analysis of the EEG signal on an event by event basis.

The purpose of the project is to continue the development of a new method of localising the regions of the brain responsible for abnormal, epileptic discharges in patients with epilepsy. The method, EEG-fMRI, combines electroencephalography (EEG) and functional magnetic resonance imaging (fMRI). EEG records the electrical activity of the brain via electrodes attached to the scalp, and is widely used in the diagnosis and study of epilepsy. FMRI uses an MRI scanner to detect changes in blood flow that occur whenever a region of the brain is active. Both techniques have a wide range of applications in the study of normal and abnormal brain function.

The advantage of combining the two techniques is that fMRI can determine very accurately the location of the active regions, while EEG has the potential to unlock their relative timing, allowing more accurate interpretation of fMRI results and hence the brain response to epileptic activity. The project will not be applied to the localisation of epileptic seizures, because of their unpredictability, the safety implications of having a seizure in an MRI scanner and problems with data quality if the patient moves during scanning. Instead, interictal discharges will be studied, which are small events occurring regularly between seizures of which the patient is unaware.

Clinically, it can be extremely important to determine which regions of the brain generate epileptic discharges, both for a better understanding of the patient's epilepsy and, in suitable cases, to help in planning surgery. EEG-fMRI is one method of doing this, and is potentially very powerful because it is very straightforward for the patient, requiring only the application of electrodes to the scalp followed by a period lying down in the MRI scanner. In addition, since most UK hospitals have MRI scanners and EEG expertise, it could take advantage of capacity that already exists within the NHS. This project will build on previous work to characterise the link between the electrical (EEG) and blood flow (fMRI) responses to interictal discharges which is necessary in order for the technique to become widely used clinically.

Patients will be recruited from epilepsy clinics after having the nature and purpose of the project explained to them. They will also receive initial screening for MRI safety to ensure that they can be scanned safely. Upon arrival at the Birmingham University Imaging Centre (BUIC), EEG electrodes will be attached to the patient's scalp using an electrode cap. When the electrodes are in place, the patient will be asked to lie down on the scanner bed. Ear plugs and head phones will be worn to minimise the noise of the scanner but this will not prevent communication with the scanner operator. Once the bed has been moved to the centre of the scanner, the patient will be asked to lie still for the remainder of the scanning session. No task will be performed. Scanning will include an anatomical image and multiple runs of fMRI. The patient will have an alarm bell which can be pressed at any time to attract the attention of the scanner operator, and fMRI scanning will have regular breaks every 5-6 minutes during which the scanner operator will talk to the patient. During the course of scanning the patient will be monitored from the control room via the window and video camera. Scanning will continue for a maximum of two hours or until the patient expresses the desire to terminate the session. The patient will then be removed from the scanner and the electrodes removed, following which the patient will be free to leave BUIC. It is expected that the whole procedure will take up to 3 hours and will be conducted once per patient, although if a patient expresses a willingness to return this may be considered.

Patients with a variety of types of epilepsy will be scanned. This will result in a comprehensive database, taking into account losses due to poor quality data, patients who wish to leave the scanner early and those in whom no interictal discharges are observed. The main result of the study will be to compare the properties of the fMRI response to the corresponding properties of each interictal discharge, in order to determine why some discharges are detectable by fMRI, while others are not. This will be done on a patient by patient basis, although comparison across patients with similar types of epilepsy will begin to help determine which patients are suitable for EEG-fMRI, a crucial question if the technique is to be applied efficiently and effectively.

Inclusion Criteria:

1. adults with no contraindications to MRI scanning. Initial screening for MR compatibility will be performed at recruitment.
2. epilepsy with interictal epileptic discharges evident on scalp EEG.
3. rarity or lack of seizures. This is an important point to minimise the risk of the patient having a seizure during scanning.
4. expected rate of interictal epileptic discharges sufficient to allow at least 10 but less than 200 events to be recorded per hour. Both too few and too many events make interpretation of the EEG-fMRI results difficult.
5. willingness to remain in the scanner for a minimum of 40 minutes. This ensures that a reasonable amount of fMRI data is collected. It is important that motivated subjects are recruited, preferably those who have previously undergone MRI scanning and so have some idea of the environment.

Exclusion Criteria:

1. contraindications to MRI scanning.
2. epilepsy with rare or absent interictal discharges, or frequent seizures.
3. patient is unable to give informed consent or understand the nature of the study.
4. neurological or psychiatric diagnosis other than epilepsy.

Patients with epilepsy with interictal epileptic discharges evident on scalp EEG.