Closing the Loop

Implanted neurostimulation systems currently on the market have seen considerable commercial and clinical success with tens of thousands of users. Nonetheless, most of the systems available today suffer from one notable shortcoming: the lack of closed-loop feedback that influences the delivery of subsequent stimulation signals.

As we report in our article on page 1, several neurotechnology firms and research institutions are now looking at closing this loop. The introduction of new stimulation leads that incorporate both stimulation and recording sites will offer new capabilities to vendors and clinicians.

Manufacturers of deep-brain stimulation systems and the neurosurgeons who implant them may be the first to benefit from this trend. As was discussed at this month’s Neural Interfaces Workshop, determining optimal stimulation locations and parameters is not always an easy task. Much of the tedium in what can often be a six- to eight-hour implantation procedure results from surgeons mapping the brain in order to find the best location for DBS leads. This process could conceivably be simplified with intraoperative recording using the same leads that will deliver therapeutic stimulation.

The flexibility these closed-loop systems would offer are important for several reasons. First, much of the brain is still very much uncharted territory and it’s conceivable that stimulation locations in the future may be different from those in use today. Second, stimulation parameters such as frequency and current level may have a profound effect on the overall success of the system. Third, there is likely to be wide variability among different patients in terms of the best locations and stimulation parameters. Often the only way to make this determination is trial and error: see what works and keep using the locations and parameters that work best and discontinue using those that don’t work.

In the case of tremor, rigidity, and other physical symptoms of movement disorders such as Parkinson’s disease, it might be reasonably easy to measure these physical attributes with implanted sensors or even surface recording devices such as EMG. In the case of neural indicators, say, for example, synchronized oscillation in the 20 to 25 Hz range from the basal ganglia, more complex recognition algorithms could be integrated into a smart programming device that collects data over time and adjusts stimulation patterns based on that data.

As new advances in neurosensing came along, the programmer could be upgraded with new software that fine-tunes the stimulation optimization process. This scenario opens the door for tighter cooperation between neurostimulation and neurosensing vendors and research teams. And closing that loop could be a healthy development for the neurotechnology industry.

James Cavuoto
Editor and Publisher