Evaluating the Efficacy of Personalised Brain Stimulation for Addiction

Current therapies for addiction largely centre on cognitive strategies and emotional management. Given that more than 66% of people will relapse after treatment, new therapeutic approaches are gravely needed. We are developing virtual brain twins to identify parameters for personalised, non-invasive deep brain stimulation targeting the neural underpinnings of addiction. To construct the virtual brain twin, fMRI data is collected while participants perform the Iowa Gambling Task, which probes reward and decision-making processes disrupted in addiction. After identifying which neural metric best predicts participant task performance, virtual brain twin parameters are optimized so that model outputs that most closely match the target neural metric. A stimulation term is added, and stimulation parameters optimized to shift the target neural metric in the direction that accords with improved task performance. Participants return for a second session of task fMRI, where they receive both personalized stimulation and generic stimulation. While data collection is ongoing, we anticipate that personalised stimulation will more effectively shift brain-behaviour relationships, setting a new course for neuromodulatory interventions for addiction.