Skip to main content

Multiscale Modelling of the Neuromuscular System for Closed Loop Deep Brain Stimulation

Periodic Reporting for period 3 - DBSModel (Multiscale Modelling of the Neuromuscular System for Closed Loop Deep Brain Stimulation)

Reporting period: 2018-08-01 to 2020-01-31

Over the past two decades, deep brain stimulation (DBS) has become established as an effective treatment for movement disorders including Parkinson’s disease, essential tremor, and dystonia. By electrically stimulating neurons deep within the brain, DBS aims to disrupt pathological, or abnormal, activity of neural circuits that promote pathological firing and are associated with the development of patient symptoms. DBS consists of high frequency electrical stimulation within affected brain regions applied via small surgically implanted electrodes. The stimulator parameters including electrical pulse strength, frequency, and duration are externally tuned within the clinic. Device re-tuning is done on a trial and error basis until an optimal parameter setting is found for each individual patient. Despite its high success in controlling movement symptoms, the exact way in which DBS works is not well understood. Furthermore, at the moment there is a strong clinical need to improve DBS methods to provide better control of symptoms for a wider range of patients, limit side-effects and extend stimulator battery life.

The project DBSmodel aims to address this need by developing an alternative ‘closed-loop’ approach for DBS that would automatically adjust stimulation parameters as needed to deliver the optimal electrical stimulation to control the patient’s symptoms at each instant in time. This type of approach offers the potential to alter the stimulation parameters, such as the strength of stimulation current, to optimise clinical benefit, minimise side effects, and reduce power consumption.
To do this, we are developing a new multiscale model of the neuromuscular system. The model will simulate neural sensing and stimulation, incorporating details of individual neurons within the brain right through to the control of muscles. The model will encompass the electric field around the electrode, the effect on individual neurons and neural networks, and the generation of muscle force.

In parallel, we are conducting experiments to measure motor neuron and muscle activity in individuals with Parkinson’s disease and in healthy volunteers. These experiments will help us to better understand the changes occurring within the nervous system in Parkinson’s disease and the way in which DBS helps to overcome them. Information from these experiments is being used to validate the computational model, and also to help identify new biomarkers of neural activity that could be used to enable continuous monitoring of patient symptoms. Using the model developed, in combination with insights gained through the experimental studies, we will design novel control strategies for closed-loop DBS. The identification of suitable closed-loop approaches for adaptive deep brain stimulation has the potential to advance the next generation of neuromodulation devices and provide more effective stimulation for patients, enabling greater control of symptoms and side-effects.
At the midterm point substantial progress has been made towards the objectives set out in DBSmodel. The development of the computational model has progressed as planned with the development of a number of models of different components within the neuromuscular system. In addition, in a number of areas, the models have extended beyond what was originally proposed, with new methods for modelling developed, opening up new avenues for research.

Computational models of the DBS electrode and networks of neurons within the brain have been developed. Using approaches similar to that used in modelling the brain, the research group have developed novel methods for simulating electric fields produced by muscle activation, incorporating for the first time detailed muscle fibre architecture based on diffusion tensor imaging. Magnetic resonance and diffusion tensor data from the hand have been acquired in a set of volunteer subjects. The acquisition of this data has allowed us to develop subject-specific electrophysiological models of the hand which incorporate anatomically accurate characteristics and the different dielectric properties of muscle, bone, fat and skin.

Substantial work has also seen the advancement of the cortico-basal ganglia network model and model of the motoneuron pool. Major achievements of the project include the development of new methods for the reduction of computational models of the complex neural dendritic tree and a new model of force production in skeletal muscle that enables us to simulate action potential propagation and excitation contraction coupling in the muscle fibre. The model is being further extended to incorporate mitochondria in skeletal muscle which will enable us to examine the role of comprised mitochondrial function. The application of engineering control strategies to the neuromuscular system has also progressed, with several novel control strategies being applied to the model.

Experimental studies which are being conducted in parallel with the computational modelling work have provided new insights into central nervous system control of muscle activity in both healthy individuals and individuals with Parkinson’s disease. Electrophysiological data have been recorded in individuals with Parkinson’s disease and healthy control subjects under a range of different conditions, and new methods for analysing these data developed. The experiments are providing insight into changes in neuromuscular control in Parkinson’s disease and are enabling us to examine the effect of therapeutic interventions. Through international collaborations we have been able to develop and test new algorithms for extracting information from surface electromyographic (EMG) data which are being used to examine changes occurring in Parkinson’s disease and also other neurological conditions, including stroke.
The primary outcome at the end of the project will be the identification of an adaptive closed-loop DBS control scheme that is capable of responding to changes in neuromuscular system properties over time, controlling both symptoms and side-effects, while minimising power consumption.

The model being developed will provide new insights into the relationship between pathological oscillatory neural activity and motor symptoms in Parkinson’s disease. New insights are being obtained into the mechanisms by which DBS exerts its therapeutic influence at the cellular, network and system levels. A number of novel methods and models have been developed to date in the project, which will contribute to progressing computational modelling of the neuromuscular system beyond the state of the art.

We are obtaining new data on motor unit activity in patients with Parkinson’s disease which will provide new information on alterations in neuromuscular control in Parkinson’s disease. As part of the foundation research for this, new insights have been obtained into the way in which the nervous system modulates motor unit activity with increasing muscle force. It is anticipated that at the end of the project, an additional outcome will be the identification of new biomarkers which could be used in early diagnosis of Parkinson’s disease or in assessing disease progression and the efficacy of different types of therapeutic interventions.