Sridevi Sarma is using computational tools to understand and fine-tune deep brain stimulation.
Insertion of an electrode during deep
brain stimulation for Parkinson's
July 2011--Deep brain stimulation, or DBS, has helped tens of thousands of patients with Parkinson’s disease. In DBS, neurosurgeons guide a fine wire electrode into brain regions involved in motor control. For many patients, this procedure dramatically reduces tremor, rigidity and other motor symptoms that characterize the disease.
But to scientists, the technique poses a mystery.
“The big question is how and why DBS works,” says Sridevi Sarma, an assistant professor of biomedical engineering who also has a faculty position with the Johns Hopkins Institute for Computational Medicine.
Sarma is seeking answers that she hopes will help demystify deep brain stimulation—and perhaps help manufacturers improve the technology, which has also been used to treat depression, chronic pain and a variety of other conditions.
Her research focuses on a region deep inside the brain called the basal ganglia. The will to move—the planning and execution of movement commences within this sprawling neural circuitry. One key part of this elaborate neural circuitry is a small area called the substantia nigra, which is rich in neurons that produce dopamine, a chemical critical to neural communication. In the healthy brain, neurons from the substantia nigra project to other parts of the basal ganglia. But in Parkinson’s disease, the substantia nigra degenerates. “So all the signals in the entire circuitry get corrupted,” says Sarma. The result: the trembling, rigid trunk and limbs, slowed movements and other symptoms seen in the disease.
Medication is generally the first line of treatment, particularly the drug levodopa, which the body converts into dopamine. But after several years many patients stop responding to drug therapy and develop serious side effects, most notably dyskinesia, or involuntary erratic movements of the limbs, trunk or face. Those patients are candidates for deep brain stimulation.
When a patient receives DBS, a one-millimeter thick-electrode is implanted into targeted areas in the basal ganglia. (The electrode is controlled by a pulse generator, a battery-controlled device similar to a heart pacemaker that is implanted under the skin just below the collarbone and connected to the electrode by an insulated wire.) The device transmits high-frequency electrical current to the brain. These electrical impulses somehow reduce the motor symptoms of Parkinson’s disease for many patients.
There are drawbacks, however. One problem is that the system’s high frequency drains the battery in the pulse generator relatively quickly. So patients must undergo a surgical procedure every few years to replace it. Another issue is that the high-frequency impulses can aggravate certain non-motor symptoms that often occur in Parkinson’s disease, such as depression, memory loss or anxiety.
“But nobody knows what another signal might do,” says Sarma. Might reducing the frequency of the DBS signal or varying its pattern in some other way produce better results with fewer side effects?
To explore these ideas, Sarma, who has a Ph.D. in electrical engineering from MIT, does not work directly with neurons. Instead, her media are mathematical models and algorithms. She collaborates with a team of physiologists at the Cleveland Clinic and University of Alabama who are using a macaque monkey model to study Parkinson’s. Sarma uses the data generated by their experiments to design and test models of the parkinsonian brain and the effects of deep brain stimulation.
Sarma’s collaborators (John Gale, at the Cleveland Clinic, and Erwin Montgomery, at the University of Alabama) recorded the electrophysiological activity of the basal ganglia of four healthy monkeys under two conditions: as the animals received deep brain stimulation and without DBS.
The scientists then gave the animals a neurotoxin that induces a Parkinson’s-like condition, and repeated the recordings, measuring basal ganglia activity while the animals received deep brain stimulation and while the animals did not receive DBS.
In analyzing the results, Sarma, with graduate student Shreya Saxena, finds significant differences between the electrical activity of the healthy brain and the brain in a Parkinson’s disease model. During rest, the basal ganglia of the healthy brain produces a steady rhythm of electrical impulses in the frequency range of 10 to 30 hertz, a pattern Sarma calls a beta rhythm. This beat, says Sarma, “seems to act like a brake; its activated when you don’t want to move.”
When an animal wants to move, Sarma and Saxena observed, the beta rhythm is suppressed and is replaced by high-frequency (30 to 100 hertz) pulses, a pattern neuroscientists call a gamma rhythm. She calls this change a “crossover” event. “The system shifts from brake to accelerate,” an event that is relayed to the muscles as a signal to move.
In the Parkinson’s disease model, however, the pattern is very different, the researchers found. During rest, many more neurons pulse in the low-frequency range—the beta rhythm. “So Parkinson’s disease has a huge brake,” says Sarma. “When a person with Parkinson’s wants to move, it’s that much harder for the brain to replace the beta rhythm with another,” which suggests, says Sarma, a much larger brake for the body to overcome to execute coordinated movements.
Deep brain stimulation, her analysis suggests, appears to overcome this hurdle by giving more basal ganglia neurons a chance to fire in the high-frequency range, that is, to cross over from the slow to the fast rhythm.
Sarma is now continuing this research by examining whether reducing the DBS current might improve the technique’s results. A lower-frequency current might put less drain on the device’s battery and cause fewer undesirable side effects.
So Sarma is creating models to test how a lower frequency DBS signal will affect the basal ganglia. She is also exploring how varying the rhythm of the input signal might affect the system. Deep brain stimulation uses a constant high-frequency current. But in the basal ganglia, neurons may employ a whole array of electrical frequencies to transmit motor information, such as how fast to move a limb and in what direction. Further research, says Sarma, could reveal how to program the DBS signal to better mimic the brain’s natural activity—which could afford patients with Parkinson’s disease even greater command of their movements.