Project title

Characterisation and automatic classification of different structural causes of epilepsy 

Supervisors


Background

1 in 100 people have epilepsy. When a child or adult is diagnosed with epilepsy they have an MRI scan as part of their assessment.  Each year in the UK, 30000 new epilepsy MRI scans are carried out. 20-30% of scans will have epilepsy-associated structural brain abnormalities. Scans are initially reported by general radiologists and 30-50% of epilepsy lesions are missed. Prognosis and treatment is dependent on the particular epilepsy aetiology. The Multi-centre Epilepsy Lesion Detection (MELD) Project has collated a dataset of over 2000 MRI scans from individuals with a variety of different structural causes of epilepsy. These include focal cortical dysplasia, hippocampal sclerosis, low-grade epilepsy associated tumours, hypothalamic hamartomas and other pathologies. The MELD Project has developed an algorithm to segment and identify these lesions. This PhD project will involve characterising the imaging features of different epilepsy pathologies and developing deep-learning tools for automatically classifying different epilepsy pathologies. These tools could eventually be incorporated into the initial radiological reporting of epilepsy patients or their presurgical evaluation.

PhD objectives and timeline

  • Year 1 – Characterisation of imaging features of 8 different structural brain abnormalities associated with epilepsy.
  • Year 2 – Develop deep learning tool to automatically classify different epilepsy lesions.
  • Year 3  – Integrate characterisation and deep learning tool into end-to-end pipeline that could be incorporated into radiological reporting of patients with epilepsy. 


Methods

The student will gain experience in python programming, AI model development and translational neuroimaging. They will develop data integration and analysis skills across imaging and clinical datasets. Finally, they will be working as part of a multidisciplinary team, collaborating with biomedical researchers, doctors and deep learning experts.

References

  1. Ripart, Mathilde, Jordan DeKraker, Maria H. Eriksson, Rory J. Piper, Siby Gopinath, Harilal Parasuram, Jiajie Mo, et al. 2024. “Automated and Interpretable Detection of Hippocampal Sclerosis in Temporal Lobe Epilepsy: AID-HS.” Annals of Neurology, November. https://doi.org/10.1002/ana.27089.
  2. Ripart, Mathilde, MELD consortium, Sophie Adler, and Konrad Wagstyl. 2024. “Multi-Pathology MRI Lesion Segmentation in a Multi-Centre Cohort of Patients with Focal Epilepsy: A MELD Study.” MIDL https://openreview.net/pdf/2a77043d316580d9ba1f320a56379abdc166c456.pdf.
  3. Ripart, Mathilde, Hannah Spitzer, Logan Z. J. Williams, Lennart Walger, Andrew Chen, Antonio Napolitano, Camilla Rossi-Espagnet, et al. 2025. “Detection of Epileptogenic Focal Cortical Dysplasia Using Graph Neural Networks: A MELD Study.” JAMA Neurology, February. https://doi.org/10.1001/jamaneurol.2024.5406.
  4. Wagstyl, Konrad, Kirstie Whitaker, Armin Raznahan, Jakob Seidlitz, Petra E. Vértes, Stephen Foldes, Zachary Humphreys, et al. 2021. “Atlas of Lesion Locations and Postsurgical Seizure Freedom in Focal Cortical Dysplasia: A MELD Study.” Epilepsia, November. https://doi.org/10.1111/epi.17130. 


Who should students contact?

Dr Konrad Wagstyl (k.wagstyl@ucl.ac.uk)or Sophie Adler (sophie.adler.13@ucl.ac.uk) 

Research topic

Paediatric surgery, Epilepsy