Dimension reduction methods, persistent homology and machine learning for EEG signal analysis of Interictal Epileptic Discharges
Annika Stiehl, Stefan Geißelsöder, Nicole Ille, Fabienne Anselstetter, Harald Bornfleth, Christian Uhl · 2025
Read Dimension reduction methods, persistent homology and machine learning for EEG signal analysis of Interictal Epileptic Discharges by Annika Stiehl, Stefan Geißelsöder, Nicole Ille, Fabienne Anselstetter, Harald Bornfleth, Christian Uhl (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-12814.txt.
Description
Annika Stiehl and co-authors build a pipeline for spotting epileptic discharges in brain wave data, in a paper published in 2025. The events, brief abnormal bursts between seizures, help diagnose epilepsy, but reading long recordings takes trained staff. Machine learning can help, yet the raw signals have far too many dimensions to feed a classifier directly.
Their chain first shrinks the data with two reduction methods, one classic and one tuned to dynamics. A shape-analysis method then studies the loops and holes in the reduced signal, a view called persistent homology. Features drawn from that view train a support vector machine to separate true discharges from background activity. The authors present the chain as a way to turn messy medical signals into compact, learnable traits.
Catalog record
| CREATOR | Annika Stiehl, Stefan Geißelsöder, Nicole Ille, Fabienne Anselstetter, Harald Bornfleth, Christian Uhl |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 8 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-12814