Stop Signal Reaction Time measured with a portable device validates optimum STN-DBS programming

Aborting an ongoing motor response (response inhibition) is controlled through a complex network, critically involving the hyper-direct pathway from supplementary motor area to the subthalamic nucleus (STN) [1]. Logan and Cowan first described the Stop Signal Reaction Time (SSRT), a measure of stopping efficiency [2]. Since then SSRT has been investigated in various neurological disorders. Parkinson’s disease (PD) patients have deficient response inhibition, which is independent of the severity of bradykinesia [3]. Levodopa and bilateral STN deep brain stimulation (DBS) both improve response inhibition in Parkinson’s patients [4,5]. Recently, our group has developed a portable device to measure reaction time and response inhibition, which is easy to use at the bedside or in the clinic. We have also introduced an improvedmeasure (optimum combination SSRT; ocSSRT) which uses a Bayesian statistical approach to enhance reproducibility [5]. In this pilot study we recruited Parkinson’s patients from our movement disorders clinic, and compared ocSSRT with standard bedside clinical assessments during optimisation of bilateral STN-DBS parameters. We included 16 patients with STN-DBS (Medtronic Activa stimulators), in whom DBS electrodes had been implanted at least six months prior to optimisation. All patients reported motor complaints with existing programming, and attended clinic to optimize simulation parameters using the N’vision 8840 physician programmer. Measurements of Movement Disorder SocietyUnified Parkinson’s disease Rating Scale Part III (MDS-UPDRS III, Motor score for PD patients) and ocSSRT were made three times: with the initial DBS setting (when the patients hadmotor symptoms), after turning off the DBS and lastly after reprogramming, when both programmer and subject felt that the settings were optimal. The ocSSRT box has a screen to display instructions and test results. A green and red LED act as the go and stop cues respectively (Fig. 1A). Patients started a trial by pressing and holding a button. They were told to release this quickly if the green LED illuminated (go trial), but to keep the button pressed if the red LED illuminated (stop trial). Responses to 192 trials were recorded (inter-trial interval 1e2.638s), in three blocks of 64 trials separated by a 60 s rest. Within each block, there were 48 go trials and 16 stop trials. The stop trials presented the red LED at four different delays (5e195 ms) after the green LED. The device measured the distribution of reaction times in the go trials, and the proportion of stop