The Neuromodulation Connectivity Dataset: Cross-Modal Data for Closed-Loop Brain Stimulation

Jasneet Kaur Dhaliwal, Ollie Turnbull

Track: Track 2: Dataset Proposal Competition

Keywords: neuromodulation, EEG, closed-loop systems, brain-computer interfaces, neuroinformatics, functional connectivity, machine learning, adaptive stimulation, neuroscience datasets

TL;DR: An EEG-based dataset linking stimulation parameters and functional connectivity to accelerate research on closed-loop neuromodulation

Abstract: Neuromodulation enables the precise modulation of human brain circuits, with promising therapeutic applications and providing a causal probe of brain function. It includes a wide range of techniques, from invasive methods such as deep brain stimulation (DBS), vagus nerve stimulation (VNS), and responsive neurostimulation (RNS), to non-invasive approaches such as transcranial focused ultrasound (tFUS), transcranial magnetic stimulation (TMS), and transcranial electrical stimulation (tES). A key shift in the field is moving from open-loop to closed-loop strategies, where stimulation is dynamically adjusted based on neural activity. However, closed-loop modulation requires accurate modelling of how differing stimulation protocols affect brain functional connectivity (FC), and how those changes in FC result in clinical or behavioural outcomes. The field is currently hindered by heterogeneous, modality-specific datasets that impede generalisable model development. We propose an expanded cross-modality dataset linking stimulation protocols, EEG activity, standardised FC readouts, and behavioural or clinical outcomes. Such a dataset would enable real-time, personalised stimulation and provide deeper circuit-level biological insight.

Submission Number: 275