Close
Supervisors: Professor Chris Clark, Professor Torsten Baldeweg
Hypothesis: neural soma imaging can detect focal cortical dysplasia and measures of neural soma density from MRI correlate with those obtained from histology.
Diffusion MRI is a quantitative imaging technique that can provide measures of tissue microstructure non-invasively. The method has been used extensively to measure damage to brain tissue in numerous neurological conditions including epilepsy in which abnormal neuronal firing or activity generates seizures. Ongoing seizures cause further damage to brain structure and/or function leading to cognitive dysfunction and other associated brain related co-morbidities. Frequently in children with focal onset epilepsy the cause or origin of the seizures is due to a lesion which may arise due to malformation in the way the cortex has developed. Focal cortical dysplasia is one such example. If a lesion can be demonstrated on MRI which is concordant with neurophysiological measures such as EEG, the patient may be offered surgery to resect the lesion. This results in seizure freedom in about 60% of patients and can be transformative in terms of quality of life and cognitive outcome. A particular difficulty in being able to offer surgery is clear evidence of a lesion on MRI. Focal cortical dysplasias can be difficult to detect on FLAIR images which generally offers the best contrast. However, focal cortical dysplasias are an example of neuronal migration disorders whereby neuronal cell bodies migrate into the white matter of the brain. In addition to these phenomena focal cortical dysplasias exhibit dysmorphic neurons with abnormal size and morphology and/or balloon cells which are unusually large cells.
We recently developed a new quantitative MRI technique based on diffusion MRI which we term neural soma imaging (1). The method models the signal in diffusion MRI as a series of tissue compartments with different characteristics. These compartments are the cell bodies, which are modelled as spherical compartments, axons and dendrites, which are modelled as anisotropic structures (with a preferred orientation) and the extracellular space. Using an advanced diffusion MRI sequence method known as B-tensor encoding the acquisition was applied in healthy volunteers providing the maps below with a spatial resolution of 2 mm isotropic voxels. The neural soma map shows strong grey/white matter contrast highlighting the density of spherical compartments, ie neuronal cell bodies in cortex and basal ganglia (see figure below). Subsequent improvements in the speed of the sequence using multi band provides an imaging time of 12 minutes which is clinically feasible.
The student will recruit 20 patients from the Department of Neurology and Neurosurgery at Great Ormond Street Hospital split into 10 lesion positive and 10 lesion negative cases based on conventional MRI including T2-weighted MRI and FLAIR. This will allow us to investigate the neural soma density of lesions which can be easily visualised on the one hand as well as providing an opportunity to investigate possible cortical abnormalities that are not visible on conventional MRI. The neural soma imaging method will be compared against FLAIR which is the current gold standard for the detection of focal cortical dysplasias. Given that neural soma imaging is based on tissue signals there will be an opportunity to modify the tissue model in an attempt to increase it’s sensitivity and specificity. For example, in the case of focal cortical dysplasia with balloon cells, much is known about the cellular dimensions of these cells, which could be included in the tissue model.
Finally, the second aim of the project is to validate the neural soma imaging method against histology. We will undertake histological analysis of tissue from 5 of the 10 cases that are lesion positive and have proceeded to surgery.
Timeline (if applicable):
Year 1 – literature review, sequence optimisation, after month 6 begin data acquisition
Year 2 – data acquisition, mathematical modelling of diffusion signal, begin histological comparison
Year 3 – optimised diffusion model, comparison with conventional methods such as FLAIR for FCD detection. Final comparison with histology. Writing-up thesis.
References:
1. Gyori, N.G., Clark, C.A., Alexander, D.C., Kaden, E. (2021). On the potential for mapping apparent neural soma density via a clinically viable diffusion MRI protocol. Neuroimage 239:118303.
2. Kerkala, L., Nery, F., Callaghan, R., Zhou, F., Gyori, N.G., Szczepankiewicz, F., Polombo, M., Parker, G.J.M., Zhang, H., Hall, M.G., Clark, C.A. (2021). Comparative analysis of signal models for microscopic fractional anisotropy estimation using q-space trajectory encoding. Neuroimage 242, 118445.
3. Gyori, N.G., Palombo, M., Clark, C.A., Zhang, H., Alexander, D.C. (2021). Training data distribution significantly impacts the estimation of tissue microstructure with machine learning. Magnetic Resonance in Medicine 87:932-947.
4. Lacerda, L.M., Clayden, J.D., Handley, S.E., Winston, G.P., Kaden, E., Tisdall, M., Cross, J.H., Liasis, A., Clark, C.A. (2020). Microstructural investigations of the visual pathways in pediatric epilepsy neurosurgery: insights from multi-shell diffusion magnetic resonance imaging. Frontiers in Neuroscience 14:269.
5. Lorio, S., Adler, S., Gunny, R., D'Arco, F., Kaden, E., Wagstyl, K., Jacques, T., Clark, C.A., Cross, J.H., Baldeweg, T., Carmichael, D. (2020). MRI profiling of focal cortical dysplasia using multi-compartment diffusion models. Epilepsia 61(3):433-444.
