The project comprised two parts. First, we assessed what function is normally carried out by the subthalamic nucleus (the area that is stimulated during DBS) during decision-making in patients with Parkinson’s disease. To this end, we recorded neural activity from the subthalamic nucleus and interconnected areas at the surface of the brain (prefrontal cortex and motor cortex) while patients were making simple decisions in which they had to judge whether a cloud of moving dots was moving to the left or right side on a computer screen. We analysed data from an existing dataset of eleven patients and a newly acquired study with eleven patients. We found that activity in the subthalamic nucleus reflected the ‘decision threshold’ determining the amount of evidence that people require before committing to a choice. Importantly, we detected two distinct correlates of such threshold adjustments; Oscillatory activity between 2-8 Hz (low frequency oscillations) was related to activity at the prefrontal cortex and predicted increased decision thresholds, but only if patients were cautious when making decision, not when they weighted speed over accuracy. Conversely, oscillations between 13-30 Hz (beta oscillations) were related to activity at the motor cortex and predicted decreased decision thresholds irrespective of patients emphasising speed or accuracy during decision-making. These findings indicate that distinct cortico-subthalamic connections are involved in determining whether people respond in haste or with caution. The results have been published in the journals Current Biology (doi: 10.1016/j.cub.2016.01.051) and eLife (doi: 10.7554/eLife.21481.). They have also been presented both as a poster and oral presentation (symposium ‘Fronto-subthalamic circuits for control of action and cognition’) at the Society for Neuroscience meeting 2016 in San Diego, USA, the world’s largest Neuroscience conference with over 30 000 attendees.
In the second part of the project, we used the same task as above, but this time we repeated the task three times for each patient; with DBS turned off, with DBS turned continuously on, and with closed loop DBS. This design allowed us to test whether DBS altered how patients controlled their responses during decision-making. We have thus far included seven patients. Preliminary analysis indicates that, in contrast to our a-priori hypothesis, closed loop DBS does affect patients’ ability to slow down responses when decision are difficult. Interestingly, this effect seems to depend strongly on the exact timing when stimulation is applied: it was only present if stimulation was applied during a short period several hundred milliseconds after the cue (cloud of moving dots) appeared on the screen, but not during other time windows. This suggests that DBS could be applied without affecting patients’ ability to slow down responses, if stimulation was turned off during the detected task-related time window. The results will be presented at an upcoming conference on DBS and optogenetics (Opto-DBS 2017) in Geneva, Switzerland. We also expect to submit the results to a peer-reviewed journal soon.
