Work

Feature extraction of motor-evoked potentials in transcranial magnetic stimulation

Year: 2024
Type: dissertation
Abstract: Transcranial magnetic stimulation (TMS) is a non-invasive brain stimulationtechnique that uses magnetic fields to stimulate nerve cells in the brain and is used as a clinical assessment tool to measure and evaluate aspects of brain function, such as motor cortex excitability, cortical mapping, and connectivity, providing insights into neurological disorders and guiding treatment decisions.TMS can activate the neurons involved in motor control resulting in motor responses at a peripheral muscle, so called motor-evoked potentials (MEPs).An in-depth understanding of MEPs in different experimental settings is essential to derive novel insights into the electrophysiology of the neuronal pathway stimulated by TMS.Therefore, this thesis focused on feature extraction and advanced analysis of MEPs.Study I analyzed MEP morphology using principal component regression (PCR).PCR was employed to extract principal components (PCs) from repeatedly measured MEP datasets.This approach effectively characterized MEP morphology via a set of PCs and determined the minimum number of trials (around 20 MEPs) needed to represent 90% of the dataset's variance.Study II investigates the maturation of neurophysiological processes from childhood to adulthood through MEP features.Participants across different age groups underwent navigated TMS, measuring MEPs in hand and forearm muscles.It shows that MEP size and duration increased with age, while latency decreased in adolescence, and most MEP features were consistent from preadolescents to adults.This research highlights evolving neurophysiological patterns and encourages further exploration with this approach with larger samples.vi Study III investigated how the stimulation coil orientation influenced a comprehensive set of MEP features, to complement previous studies, which only concentrated on the MEP amplitude.It proved that the coil orientation strongly affected the MEP size, while its effect was minor in MEP timing features.In conclusion, these studies demonstrate that the extracted MEP features are potentially neurophysiological indicators in TMS-related research, provide a ground knowledge to study MEP features further in clinical assessments of motor disorders and motor development.Furthermore, PCR was proved as a robust technique in MEP pre-processing and analysis as it improved the interpretation of MEP and the extraction of MEP features.The developmental model of MEP features from Study II depicts neurophysiological changes of MEP features during human motor development, bringing novel insights into the motor-related neurophysiology across developmental stages.Study III illustrates the effect of coil stimulation positioning on the corresponding elicited MEP.These studies also enabled the author to construct MEPFeatX, a MATLAB package for automated extraction of MEP features in various TMS protocols.It offers comprehensive MEP analysis across various paradigms and easy integration into analysis pipelines.Thus, MEPFeatX enables the use of the methodology presented in this thesis via an open-source software package.
Language: English
Subfield: Neurology
Open Access status: green
OpenAlex ID: W7215167982
Cites: 168
Cited by:
Related to: 10