Machine Learning Based Investigation of the Imaging and Genetic Profile of Drug-Resistant Epilepsy
Lead Research Organisation:
University College London
Department Name: Medical Physics and Biomedical Eng
Abstract
Brief description
Epilepsy is a group of neurological conditions that share the common characteristic of epileptic seizures. There are many types of epilepsy and many types of seizure. The diagnosis of epilepsy typically follows the occurrence of two or more seizures. While anyone can develop epilepsy at any point in life, it is most commonly diagnosed in children and people over 65 years of age. Currently, about 60 million people are suffering from epilepsy worldwide and over 500,000 in the UK alone.
Anti-epileptic drugs (AED) are intended to reduce the frequency of seizures or even completely eliminate them. If AED treatment works, it allows people living with epilepsy to lead normal lives. However, in about one third of people none of the available drugs or combinations of drugs stop the seizures. These cases are considered treatment-resistant epilepsies and the underlying biology is still poorly understood.
Aims and objectives
In this project, we will help to advance knowledge about drug-resistant epilepsy. Using a large database of neuroimaging data from people with epilepsy, the first aim is to establish the imaging signature of drug-resistant epilepsy compared to drug-responsive epilepsy. That is, which brain regions show different cortical thickness or structural and functional connectivity changes in resistant vs responsive epilepsy? The second phase of the project involves an imaging-genetics approach to elucidate the genetic origin of drug-resistant epilepsy, which may help to identify drug targets for currently drug-resistant epilepsy.
Novelty of the research methodology
In this research project we will combine high-dimensional imaging data with high-dimensional genetics data (e.g., whole genome genotyping) in thousands of subjects. To make the research feasible this will require to adapt and modify existing machine learning methods and, more importantly, work towards methodological solutions that will enable us to leverage existing knowledge about genome organisation and molecular biology.
Alignment with EPSRC's strategies and research areas
This project is aligned with 'Healthcare technologies' grand challenge. In particular with 'optimising treatments'. The methodological contribution will be on data analytic methods to identify disease phenotypes.
Companies and collaborators involved
UCB Pharma (https://www.ucb.com/)
Prof. Sanjay Sisodiya at UCL (https://iris.ucl.ac.uk/iris/browse/profile?upi=SMSIS90)
Epilepsy is a group of neurological conditions that share the common characteristic of epileptic seizures. There are many types of epilepsy and many types of seizure. The diagnosis of epilepsy typically follows the occurrence of two or more seizures. While anyone can develop epilepsy at any point in life, it is most commonly diagnosed in children and people over 65 years of age. Currently, about 60 million people are suffering from epilepsy worldwide and over 500,000 in the UK alone.
Anti-epileptic drugs (AED) are intended to reduce the frequency of seizures or even completely eliminate them. If AED treatment works, it allows people living with epilepsy to lead normal lives. However, in about one third of people none of the available drugs or combinations of drugs stop the seizures. These cases are considered treatment-resistant epilepsies and the underlying biology is still poorly understood.
Aims and objectives
In this project, we will help to advance knowledge about drug-resistant epilepsy. Using a large database of neuroimaging data from people with epilepsy, the first aim is to establish the imaging signature of drug-resistant epilepsy compared to drug-responsive epilepsy. That is, which brain regions show different cortical thickness or structural and functional connectivity changes in resistant vs responsive epilepsy? The second phase of the project involves an imaging-genetics approach to elucidate the genetic origin of drug-resistant epilepsy, which may help to identify drug targets for currently drug-resistant epilepsy.
Novelty of the research methodology
In this research project we will combine high-dimensional imaging data with high-dimensional genetics data (e.g., whole genome genotyping) in thousands of subjects. To make the research feasible this will require to adapt and modify existing machine learning methods and, more importantly, work towards methodological solutions that will enable us to leverage existing knowledge about genome organisation and molecular biology.
Alignment with EPSRC's strategies and research areas
This project is aligned with 'Healthcare technologies' grand challenge. In particular with 'optimising treatments'. The methodological contribution will be on data analytic methods to identify disease phenotypes.
Companies and collaborators involved
UCB Pharma (https://www.ucb.com/)
Prof. Sanjay Sisodiya at UCL (https://iris.ucl.ac.uk/iris/browse/profile?upi=SMSIS90)
Organisations
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/R513143/1 | 01/10/2018 | 30/09/2023 | |||
| 2243537 | Studentship | EP/R513143/1 | 15/10/2018 | 14/10/2022 | Seymour Lopez |