gms | German Medical Science

Objective: Deep brain stimulation (DBS) is a surgical procedure used, inter alia, for treatment of different tremor syndromes. Therapeutic success strongly depends on correct positioning of the implanted electrodes and the choice of the stimulation parameters. Therefore, significant effort is invested during surgery in finding the optimal settings. The surgery is usually performed with the patient being awake or under analgosedation and performing simple exercises, so that his or her movements can be assessed. To facilitate the process we have implemented a simple system for intra-operative quantification and documentation of tremor. Here we show its validity and reliability.

Method: An accelerometer-equipped Android smartphone was used as the sensor for measuring tremor. The accelerometer sensitivity is about 0.05 m/s² and the sampling frequency about 40 Hz (irregular). The smartphone is bound by an elastic strap to the patient’s wrist and continuously transmits data to a laptop computer. The laptop processes and visualizes the data both in time and frequency domains. Selected data and information about electrode settings are stored for further analysis. To validate the measurement results, we have correlated a) the measured tremor frequency with and without stimulation (the frequency should not change) and b) the measured dominant tremor amplitude with the neurologist's tremor assessment on the scale 0–4.

Results: We have tested the system in 13 surgeries (5 essential tremor, 5 Parkinson, 2 multiple sclerosis, 1 Holmes tremor). The targets were VIM (8 cases) and STN. For five patients we measured the tremor on both body sides, and on only one for the remaining patients. Two patients were taken out of the study because no tremor could be detected during surgery. In all, 32 measurements were analyzed. A correlation (R² = 0.45, F = 12.35, line slope = 0.95) has been found between tremor frequencies with and without stimulation. The logarithm of the measured amplitude and the neurologist's intra-operative assessment have been shown to correlate with R² = 0.6 and F = 44.72. The amplitude drop caused by stimulation has been 0.43 ± 0.55 m/s².

Conclusions: It has been shown that the stimulation does not influence tremor frequency. A trend in amplitude reduction has also been shown. However, the described system does not currently qualify as a medical device, so that its application is limited to documentation and providing informative feedback to the surgeon.