Brain Imaging

Impact case study (from the UCL Impact website)

Novel MR imaging methods

P Bartlett, J Burdett, JS Duncan, S Eriksson, N Focke, R Samson, MR Symms
Collaborator: GJ Barker

We have developed novel ways to obtain and analyse MRI scans to identify subtle abnormalities of the brain, that may be both causes and consequences of epilepsy.

We have implemented a method to examine the signal intensity on FLAIR scans on a voxel-basis, and validated this with an evaluation of focal cortical dysplasia, followed by the use of the technique to examine individuals in whom no abnormality had been identified. Using a voxel-based analysis of diffusion tensor imaging scans, we identified the extent of abnormalities of the grey matter, ipsilaterally and contralaterally in temporal lobe epilepsy. We are working on the correction of B1 and B0 variations, in the 3T scanner to enable reproducible quantitative magnetisation transfer imaging (QMT) and T1 mapping. We are implementing the methods for aligning pathological samples from surgical resections with in vivo MRI and using this to enable precise correlation of MRI signal with the underlying pathology.

Visualising the connections in the brain

JS Duncan, MJ Koepp, R Powell, MR Symms, C Vollmar, S Vulliemoz, M Yogarajah
Collaborators: GJ Barker, D Alexander, GJ Parker

We are mapping the white matter pathways in the brain that connect vital parts of the brain for vision, language and memory, and seeing how these are affected in different epilepsies.

We are collaborating with Geoff Parker, now in Manchester, to use Pico tractography to visualise the connections of the parts of the brain involved in language, memory and vision. Having established the normal patterns of the white matter in healthy subjects, we showed that in temporal lobe epilepsy, the connections of Broca’s area were reduced on the side of a temporal lobe focus and increased on the other side of the brain. Of clinical relevance, the stronger the language connections on the side of the brain that is subject to temporal lobe resection, the more likely it is for there to be language difficulties after the operation.

In individuals with left temporal lobe epilepsy, the connections of the parahippocampal gyrus on the left side are reduced, which goes some way to explaining the memory difficulties that commonly accompany temporal lobe epilepsy. We are now mapping the parahippocampal gyrus, the arcuate and the uncinate fasciculi before and after temporal lobe resections to understand the effects the surgery has on the organisation of the brain.

We have implemented the methods for visualising the optic radiation in vivo, and have acquired these data prior to temporal lobe resection, and are validating methods to use these data to predict the risk of a visual field defect as a result of surgery.

We are now beginning to apply these methods to the frontal lobe and frontal lobe functions, to determine the anatomy of the connections of these eloquent areas.

Imaging the functions of the brain

S Bonelli, JS Duncan, MJ Koepp, R Powell, MR Symms, M Yogarajah

We are identifying the parts of the brain that are involved in language and memory, and using these results to predict and reduce the risks of epilepsy surgery.

We have acquired functional MRI studies of memory for pictures, faces and words in a large cohort of patients with left and right temporal lobe epilepsy, before and after temporal lobe resection and are now analysing these data to determine the utility of these methods to predict the change in memory that may occur as a result of the surgery.

Activation of the amygdala occurs on viewing material with an emotional content and we have found that this activation is predictive of the level of anxiety and depression that may occur following surgery. If this finding is confirmed in larger series, this will enable more precise targeting of support to vulnerable individuals.
We have just started to examine frontal lobe functions using a range of fMRI paradigms, and will be examining the functions of the frontal lobes and how these are affected in frontal lobe, temporal lobe and generalised epilepsies.

Positron emission tomography

MJ Koepp, JS Duncan

We are using PET tracers to investigate the chemical abnormalities and their location in the brain, which are involved in epilepsy and epileptic seizures.

We have set up an international network of collaborating PET centres in Manchester, Amsterdam and Vienna to co-ordinate activities to develop in-vivo biomarker for drug-resistance. In collaboration with the group in Amsterdam, we published the first human study using [11C]-verapamil as radiotracer for P-gp function in patients with drug-resistant epilepsy (Epilepsia 2007). This work continues with the development of better suited in-vivo PET biomarker for multidrug transporter function as a generic tool for the prediction, diagnosis, monitoring and prognosis of major CNS diseases, supported by an FP-7 multi-national large-scale collaborative programme grant. An in-vivo imaging biomarker of multidrug transporter function is essential for identifying altered transporter activity in individual patients. Such a biomarker will provide the means for predicting treatment response in individual patients, which is the necessary first step for the development of more effective treatment strategies.

Imaging of paroxysmal brain activity using EEG-correlated fMRI in patients with epilepsy

L Lemieux, JS Duncan, H Laufs (Frankfurt), D Carmichael, K Hamandi, R Rodionov, P Allen, N Toms, G Bahlke, R Thornton, A McEvoy, MC Walker, SJ Smith, S Cannadathu, A Vaudano, S Vulliemoz.

This work aims to improve our ability to localise the epileptic focus and our understanding of other epileptiform activity by recording electrical brainwaves and performing MR brain scans at the same time.

Our activities have continued to focus on the recruitment and scanning of patients from NHNN and three collaborating centres (King’s College Hospital, Hôpital La Timone (Marseille) and the Frenchay Hospital in Bristol) as part of a 5-year MRC-funded programme. The aim of the 5-year programme is to determine whether non-invasive MRI methods may provide the necessary information and obviate the requirement for invasive EEG recordings. Among important methodological developments we have developed a new way of revealing patterns apparently linked to epilepsy in images which could have dramatic implications for the future of fMRI in the study of patients with epilepsy. Furthermore, we have started looking at connectivity patterns linked to paroxysmal activity, both in focal and generalised epilepsy. Simulations and experimental work on the safety of MR scanning in the presence of intracranial electrodes has produced extremely promising results.

Serial quantitative MR in patients with epilepsy

L Lemieux, M Chupin (Paris), J Burdett

This work aims to improve our ability to detect and quantify small changes in the brain linked to disease progression.

This project focuses on the refinement of automated segmentation of the hippocampus and amygdala in serial high-resolution brain scans to improve our ability to perform very long-term longitudinal studies. We have now implemented a fully automatic software method that allows us to find, outline and measure the hippocampi and amygdale in our MRI scans. We are now embarking on the evaluation of the method’s performance in data from patients who were scanned over long periods of time, the development of new ways to perform volume measurements in repeated scans and the implementation of the technique in clinical routine. We envisage that this development will lead to much more reliable and sensitive measures of brain damage in relation to disease progression.

Correlation of imaging and pathology

S.Sisodiya, M Thom, S Eriksson, J Duncan

Inevitably, detailed microscopic neuropathological examination of human brain tissue in epilepsy is only rarely possible; such examination remains of great value. We have been developing detailed methods to more clearly relate findings from new, high-resolution quantitative magnetic imaging resonance methods and advanced quantitative neuropathological techniques that examine neuronal parameters (such as size, density, gene expression). We were able to show that some imaging measures appear to reflect direct tissue properties. On the other hand, some previous assumed relationships, commonly accepted in the research community, would appear to have less sound biological bases. Further correlative work, including imaging at high field strengths, continues to be developed.

Page last modified on 17 dec 14 13:31