Imaging Large Scale Networks in Epilepsy

Epilepsy is a frequent neurological disease that affects up to 1% of the population in most of the Western countries, with devastating consequences, in particular for younger people. Due to recurrent impaired consciousness or falls, the patients may encounter learning difficulties, remain excluded from the job market and even encounter an increased mortality. Thus, if we better understand from where the epilepsy comes, we might be able to offer more targeted treatment possibilities, be it with surgery or neurostimulation of selected structures.

In the context of the grant, we further improved the localizing properties of multichannel EEG (with up to 256 scalp electrodes, electric source imaging [ESI]), functional MRI and so-called diffusion-tensor imaging (DTI), which aims at visualizing the fibers connecting one brain structure with the others. Thanks to a large data base obtained during the first phase of the grant, we could show

1. that ESI has an excellent localizing value in patients admitted for possible surgery, and
2. that the fMRI shows a network which resembles the one found with intracranial EEG or ESI, but with a much higher precision. DTI also turned out to be promising.

However, in order to fully use this technique, we needed a data base of healthy controls, which is now accomplished. Our first observations confirm that DTI/DSI is capable to show physiological or aberrant fiber bundles in patients and this can be reliably done, with similar results, in all MR machines of the three centers (University Hospitals of Geneva, Lausanne, and Berne).

While we focused very much on epileptogenic abnormalities when there is no clinical seizure (called the interictal state), we will now extend our studies to visualize the seizure activity and determine if the spread follows what is suggested by the interictal ESI, fMRI or DTI findings. If this is true, we might not need to see the patient’s seizure to know where the epilepsy comes from or which epilepsy disease the patient suffers from, but just record a short piece of EEG.

We will also probe the increased value to very performing MRIs (7T) which show much more precisely the brain anatomy. Our studies will not only include patients who are worked up for possible surgery and/or adults, but also patients with other forms of epilepsy, including young children seen mainly by our neuropediatric colleagues. Furthermore, we like to know what happens if the brain is at “rest” in our patients. In healthy subjects, distinct areas are activated (or inactivated) and supposedly this is a sign of mental health. While we do not expect a major anomaly in this resting network, there might be subtle changes which explain attention or memory difficulties in our patients.

Finally, we make use of those patients who have to undergo brain surgery and find out which components in the MRI relate to tissue composition in small structures in the mm range. Any result from this analysis will be extremely helpful for other patients who suffer different neurological diseases (dementia, multiple sclerosis) and usually have no brain surgery. The mouse model, developed by the fourth group, resembles very much what is found in patients and will help us to better understand epilepsy on the cellular level. Like in our previous grant, most of our studies aim to explain the neuronal activity at the individual level, which is required for a tailored care, albeit we will also try to find “laws” in larger patient groups vs control subjects. With our common efforts, we hope to substantially decrease the number of patients who do not respond to drug treatment, which is currently still around 30%.