From May 9th to 14th, 2026, the International Neuromodulation Society (INS) held its congress in Lisbon, Portugal, under the theme "Bridging Minds, Machines, and Medicine." The event brought together more than 1,800 professionals from over 70 countries — clinicians, engineers, and researchers working together to advance neuromodulation across the world. Across six days, sessions explored how emerging technologies, from novel stimulation hardware to artificial intelligence, are reshaping the treatment of neurological disorders.

The congress gathered leading experts spanning neurology, neurosurgery, biomedical engineering, and computer science, alongside PhD and Master students drawn to the frontiers of brain–machine interfaces and data-driven medicine. Participants examined the latest clinical trial outcomes, refined stimulation paradigms, and the increasingly central role of machine learning in personalizing therapy and predicting patient response.

One highlight, presented as both a talk and a poster, addressed a pressing clinical challenge: predicting which epilepsy patients are most likely to respond to neurostimulation. Drawing on data from the EASEE II and PIMIDES studies, the work combined questionnaire-based clinical variables with EEG-derived spectral features to train a suite of machine-learning classifiers. Using variational-autoencoder latent representations of EEG recordings, the best-performing model reached a test accuracy of roughly 75 % in identifying treatment responders among drug-resistant patients. The results demonstrate how multimodal data integration can sharpen therapeutic precision and support more personalized neuromodulation — pointing toward earlier patient selection, reduced trial-and-error, and lower healthcare costs as the approach is validated in larger prospective cohorts.

Author: Marcus Buchwald

Date: 10/06/2026