UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Optimisation of a Wearable Neuromodulator for Migraine Using Computational Methods

Salkim, Enver; (2019) Optimisation of a Wearable Neuromodulator for Migraine Using Computational Methods. Doctoral thesis (Ph.D), UCL (University College London). Green open access

[thumbnail of Salkim_10067198_thesis.pdf]

Download (6MB) | Preview


Migraine is the third most common neurological disorder and the sixth cause of disability. It may be characterized by a headache, nausea, vomiting, photo- phobia and phonophobia. Available pharmaceutical treatments of migraine are not completely effective and have troublesome side-effects. Thus, there is a need for alternative treatments such as neuromodulation. Neuromodulation may be delivered invasively; however, this exposes the patients to the associated risks. Transcutaneous electrical nerve stimulation is a non-invasive technique that is widely used to relieve pain. A significant number of migraine sufferers complaint the symptoms of pain originating in the frontal region of the head. Thus, mi- graine may be associated with the supraorbital nerve and supratrochlear nerve which passes below the frontal bone exits from the orbital rim and penetrates the corrugator and frontalis muscles. Transcutaneous frontal nerve stimulation has been applied on a large group of patients who have episodic migraine us- ing a device called Cefaly. This study produced mixed results (50% response rate). A post–marketing survey led to 53% satisfaction while the most limiting factor is reported to be paraesthesia and painful sensation. The possible causes of these inconclusive results may be associated with neuroanatomical variations, patient compliance and neurophysiological effects. The most plausible cause may be related to the neuroanatomical variations across different subjects. The neu- roanatomical variations may lead to excessively high current levels being required. Since this solution is patient–operated, these relatively high required levels are not applied. In addition, as the electrodes are positioned near pain–sensitive structures, pain may be induced even at low current levels, further limiting the efficacy of the solution. There has been no robust investigation identifying the underlying causes of ineffi- cacy. This is partly due to the physical limitations of studying the neuroanatomy of each subject and different settings of electrode arrangements. Computational models may enable researchers to estimate current stimulation thresholds in neu- romodulation therapy and investigate the effects of various parameters. Such computational models are composed of a volume conductor model and an ad- vanced Hodgkin–Huxley–type model of neural tissue referred to as a hybrid model. Once the human head anatomy, the human nervous system and available solu- tions for migraine are detailed, the computational model of the human head is generated. A highly detailed human head model based on magnetic resonance imaging (MRI) studies, microscopic structure of the skin(including sweat ducts, keratinocytes and lipid) and those of a simplified head model (which built from geometric shapes) are compared based on neural excitation to assess the usabil- ity of geometrically realistic(simplified) human head models in the subsequent studies to save computations cost. The induced electric field due to an electrode setting is simulated in the volume conductor model and the resulting electric potential values along the nerve are passed on to the neural model to simulate nerve’s response. It is shown that a simplified model may be used with a marginal error (≈2%) in the subsequent work when assessing the effect of neuroanatomical variations on the efficacy of the target solution and possible ensuing optimiza- tions. The first step is to identify if neuroanatomical variations had any effect on the required stimulus current levels using state of the art computational bio–models. Ten realistic human head models are developed by varying thirteen neuroanatom- ical features including human head size, thicknesses of the tissue layers and vari- ations in the courses of the nerve by considering their respective statistical distributions as reported in the literature. A novel algorithm is developed to account for the variations of the nerve in different individuals and mimic statistically relevant large population. In each case, the required stimulus current levels are simulated. The findings show that the combined neuroanatomical variations have a significant effect on the neural response for the electrode setting used in Cefaly device. Therefore, a potential improvement is to align the axis of electrodes with the target nerve, so that the electrical potential along the trajectory of the nerve changes polarity. This may lead to lower required stimulus current levels. Align- ing electrodes with the nerve, the required current may be reduced by at least 60%. This new orientation reduces current density near pain– sensitive struc- tures by diverting the current away from them, which may lead to a higher level of patient compliance, further improving the efficacy of the solution. Using an electrodes array arrangement, the required current levels is further reduced due to incorporating multiple electrodes array elements to maximise the variations of the electrical field in the simulation of the fibres in one phase. The findings of this thesis indicate that the highly detailed human head model can be simplified while minimally affecting the outcome. Additionally, it is shown that neuroanatomical variations have a significant impact on the stimulus current thresholds but it is not possible to conclude if these thresholds solely depend on a specific neuroanatomical variation. The relatively high required levels of the stimulus currents are beyond the current capabilities of existing device and pos- sible pain thresholds. Furthermore, the proposed new electrode arrangement has multiple benefits including the reduction of the stimulus current levels and diver- sion of current spread from possible pain–sensitive structures. This improvement, based on modelling, can potentially improve the clinical outcome of the neuro- modulator substantially if confirmed in the subsequent clinical studies.

Type: Thesis (Doctoral)
Qualification: Ph.D
Title: Optimisation of a Wearable Neuromodulator for Migraine Using Computational Methods
Event: UCL (University College London)
Open access status: An open access version is available from UCL Discovery
Language: English
Additional information: Copyright © The Author 2019. Original content in this thesis is licensed under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0) Licence (https://creativecommons.org/licenses/by/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request.
UCL classification: UCL
UCL > Provost and Vice Provost Offices > UCL BEAMS
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Electronic and Electrical Eng
URI: https://discovery.ucl.ac.uk/id/eprint/10067198
Downloads since deposit
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item