Author: Azinsadat Janani
Janani, Azinsadat, 2019 Quantitating tonic muscle activity in head and neck for artefact removal or disease understanding, Flinders University, College of Science and Engineering
Terms of Use: This electronic version is (or will be) made publicly available by Flinders University in accordance with its open access policy for student theses. Copyright in this thesis remains with the author. You may use this material for uses permitted under the Copyright Act 1968. If you are the owner of any included third party copyright material and/or you believe that any material has been made available without permission of the copyright owner please contact copyright@flinders.edu.au with the details.
Scalp electrical recordings using surface electrodes are traditionally used to record brain signals (EEG) but such recordings contain other biological signals such as cranial and upper cervical muscle signals (EMG), cardiac signals (ECG), etc.
Recent studies have shown that even during the “relaxed” condition, sitting or reclining, many cranial and upper cervical muscles are involuntarily contracted to maintain posture, facial expression, etc. Their activity has a broad-band spectrum that overlaps the spectrum of brain and exceeds it in power. Hence, the effect of tonic muscle activity during a usual scalp electrical recording is too large to be ignored.
On the one hand, separating and removing (pruning) these tonic muscle signals from scalp electrical recordings is an issue in brain studies. On the other hand, separating and keeping these tonic muscle signals (quantitating) is valuable for treatment and/or understanding of the role of muscle in some medical conditions, such as headache.
In this thesis, using the unique database of pharmacologically induced paralysis subjects, I evaluate the effectiveness of some current advanced signal processing algorithms (blind source separation) in the automated reduction of tonic cranial and upper cervical muscle activity from scalp electrical recordings. I then study one poorly-performing algorithm (canonical correlation analysis) in detail, and propose an extension with improved results. I also propose a completely new approach to muscle pruning, based on source localisation. Acknowledging the difference in approach between these algorithms, I explore the complementary effect of double pruning approaches targeting different features of muscle signals, and show that tonic muscle reduction using double pruning approaches is significantly more effective than single pruning approaches.
I also describe an “inverted” use of muscle pruning algorithms, and propose a new holistic cranial and upper cervical muscle quantitation approach using a high-density EEG cap. This approach is validated using scalp electrical recordings from subjects suffering from diseases associating with increased muscle tension. Applying this approach on scalp electrical recordings of migraineurs and controls reveals that there is more cranial and upper cervical muscle activity in migraineurs than controls. This result diminishes one of the conceptual distinctions between migraine and tension-type headache.
Keywords: Scalp electrical recording. brain activity, muscle activity, headache, blind source seperation
Subject: Biotechnology thesis
Thesis type: Doctor of Philosophy
Completed: 2019
School: College of Science and Engineering
Supervisor: Associate Professor Kenneth Pope