

Petr Klimes
25 posts









🚨 New paper alert! 🚨 Our framework provides a comprehensive and reliable method for evaluating ML models in epileptogenic zone localization, enhancing precision and clinical relevance. Advancing AI in epilepsy care! 💡🧠

Key point: High-frequency features do not provide any additional value to the results of standard low-frequency #EEG analysis. doi.org/10.1111/epi.18… #epilepsy #ilae #HFOs #interictalepileptoformdischarges #machinelearning @IlaeWeb @epilepsiajourn @WileyNeuro


Key point: High-frequency features do not provide any additional value to the results of standard low-frequency #EEG analysis. doi.org/10.1111/epi.18… #epilepsy #ilae #HFOs #interictalepileptoformdischarges #machinelearning @IlaeWeb @epilepsiajourn @WileyNeuro




New Research: Case report: Insulinoma masquerades as epilepsy – quantitative EEG analysis: Insulinomas are rare gastrointestinal tumors with an incidence of 1–3 per million inhabitants annually. These tumors result in excessive insulin… frontiersin.org/articles/10.33… #Neurology

[Epilepsy] Chybowski et al.: "Short segment of iEEG can achieve similar results in localization of the epileptogenic zone as a model based on long recordings." doi.org/10.1016/j.clin…









Machine-designed neural network architectures outperform heuristic, human-made ones. More in our new paper about "Genetic algorithm optimization for improved reliability in iEEG analysis" published last week iopscience.iop.org/article/10.108…