The use of artificial Intelligence in the diagnosis of epilepsy v1
Research type
Research Study
Full title
Smart Epilepsy Diagnosis Tool
IRAS ID
327649
Contact name
John Watson McMaster
Contact email
Sponsor organisation
Quarriers
Duration of Study in the UK
1 years, 0 months, 0 days
Research summary
The William Quarrier Scottish Epilepsy Centre is a specialist diagnoses centre based in Glasgow. Currently, patients will be admitted, and medication withdrawn in a controlled environment and will be put onto EEG recording in the hope that a seizure will take place. If there is a recording of 24 hours, then this can take almost 24 hours to review the recording, or the physiologist will speed through the recording with the risk of missing a seizure. This EEG recorded information will allow the clinician to review and diagnose what type of seizure the patient has had, then the Dr can prescribe the best medication. We in Quarriers in partnership with GCU and Datalab intend to use Artificial Intelligence to read the EEG Recording. We have been running a feasibility pilot using publicly available data and the results have been better than predicted. The intention is to train the AI model using the historical data that Quarriers have captured over the last 12 years. The better the data the more accurate the outcomes will be. The AI algorithm will be able to identify any area on the 24-hour recording that it believes need further investigation for a clinician to review. The AI Model will take a nano- second to review the 24 hour recording and will be more accurate because this eliminates the human error element.
REC name
London - Camberwell St Giles Research Ethics Committee
REC reference
23/LO/0760
Date of REC Opinion
28 Sep 2023
REC opinion
Unfavourable Opinion