Studying the Role of the Basal Ganglia in Motor Symptoms Using Deep Brain Stimulation

This study investigates how the basal ganglia contribute to motor symptoms like tremor, bradykinesia and muscle spasm. The basic research approach is to record from sites in the basal ganglia whilst patients are symptomatic, so that brain waves can be correlated with symptoms/signs. Once a brain wave is implicated in an aspect of abnormal movement, the researchers can try and confirm its central role in function or dysfunction by triggering stimulation whenever the brain wave is picked up. For stimulation the researchers will use the same high frequency stimulation (130 Hz) as used clinically, as this is thought to effectively suppress neural activity at the stimulation site. Thus, if a given brain wave is important in, for example, slowing movement, then by triggering stimulation whenever this brain wave is big it can be expected that movement speed will be increased.

The investigators hope to follow this two-stage procedure to document the role of the different brain activities picked up from basal ganglia sites in driving tremor, muscle spasm and slowness of movement in patients with Parkinson's disease, dystonia and essential tremor. This study is important, as if the researchers can alter brain function and specific symptoms with stimulation they can use the same form of feedback-controlled stimulation as a potentially efficient form of treatment. Conventional deep brain stimulation delivers fixed stimulation all of the time. For example, researchers are beginning to see that stimulation control based on feedback from beta activity in the basal ganglia may have advantages over conventional continuous deep brain stimulation in treating Parkinson's disease.

The current study is particularly interested in the processes contributing to slowness (bradykinesia) and rigidity (stiffness) in patients with Parkinson's disease, tremor in patients with Parkinson's disease and Essential Tremor, and muscle spasms in patients with dystonia.

1. Bradykinesia and Rigidity in patients with Parkinson's disease Here there is already evidence that these impairments are associated with beta frequency band activity (~20Hz). Such activity is exaggerated in patients with Parkinson's disease in whom it comes in bursts lasting several hundred milliseconds or even longer. The investigators have already shown that by triggering stimulation when bursts of beta activity occur they can speed up movement and reduce rigidity. In the present study they are interested in (a) determining whether it is necessary to trigger off beta bursts or whether it is sufficient to trigger off the general level of beta activity (ie averaged over long periods), (b) whether if it is necessary to trigger off all beta bursts, or is it just the long bursts that need to be triggered off, and (c) whether triggered stimulation is also sufficient to control tremor where this is a co-existent symptom. Exploration of these issues requires the investigators to record basal ganglia activity (the feedback) and to deliver stimulation, whilst varying how the feedback is processed before driving the stimulation. In engineering terms the investigators vary the signal processing and control policy details, but the net result is feedback-controlled deep brain stimulation. Note that the investigators only control the amplitude of stimulation within a clinically determined range that goes no higher than the threshold for eliciting side-effects. All the remaining stimulation parameters, e.g. frequency and pulse width, are set to standard clinical settings.
2. Tremor in patients with Parkinson's disease or Essential Tremor Here the evidence that tremor is associated with a discrete brain activity is less robust, although oscillations at tremor frequency (and twice this) are suspected of playing a role. In conditions where the investigators are unsure of the exact nature of the factors contributing to a state, in this case tremor, they often use machine learning to find the relevant factors. Here the investigators propose to record both basal ganglia activity and tremor in the limbs and then use these with machine learning algorithms to point out the relevant combination of signals associated with tremor. The investigators can then use machine learning outputs to tell them how to control tremor with stimulation, whilst interrogating the weights of the inputs to the machine learning algorithms to deduce the important relationships. As above, they will explore the optimal signal processing and control policy details, but the net result is feedback-controlled deep brain stimulation. Note that they only control the amplitude of stimulation within a clinically determined range that goes no higher than the threshold for eliciting side-effects. All the remaining stimulation parameters, e.g. frequency and pulse width, are set to standard clinical settings.
3. Involuntary muscle spasms in patients with Dystonia Here the evidence that muscle spasms are associated with a discrete brain activity is also relatively weak, although oscillations at theta-alpha frequencies (5-12 Hz) are suspected of playing a role. The investigators propose to record both basal ganglia activity and muscle spasms in the body and then use these with machine learning algorithms to point out the relevant combination of signals associated with muscle spasms. They can then use machine learning outputs to tell them how to control muscle spasms with stimulation. As above, the investigators will explore the optimal signal processing and control policy details, but the net result is feedback-controlled deep brain stimulation. Note that they only control the amplitude of stimulation within a clinically determined range that goes no higher than the threshold for eliciting side-effects. All the remaining stimulation parameters, e.g. frequency and pulse width, are set to standard clinical settings.

Techniques to be used