Detection of Interictal Discharges With Convolutional Neural Networks Using Discrete Ordered Multichannel Intracranial EEG
- Submitting institution
-
The University of Surrey
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 9019109_1
- Type
- D - Journal article
- DOI
-
10.1109/TNSRE.2017.2755770
- Title of journal
- IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Article number
- -
- First page
- 2285
- Volume
- 25
- Issue
- 12
- ISSN
- 1534-4320
- Open access status
- Compliant
- Month of publication
- September
- Year of publication
- 2017
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
-
- Research group(s)
-
-
- Citation count
- 22
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Less than 9% of interictal epileptiform discharges (IED) are visible in scalp EEG, yet our technique detects these spikes at an accuracy of 89%. This is a major breakthrough for two reasons. First, the high reliability of our method provides a key tool for clinicians to save hours of analysis and preliminary surgical interventions. Second, it is the first work in deep learning for *biomedicine*, which shed light on *why* such high accuracy is possible. We showed that the learning parameters converge towards the morphology of IEDs. Our work has been followed by experts in epilepsy, neuroscience, and machine learning.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -