This thesis presents methods for detecting consciousness in patients with complete locked-in syndrome (CLIS). CLIS patients are unable to speak and have lost all muscle movement. Externally, the internal brain activity of such patients cannot be easily perceived, but CLIS patients are considered to be still conscious and cognitively active. Detecting the current state of consciousness of CLIS patients is non-trivial, and it is difficult to ascertain whether CLIS patients are conscious or not. Thus, it is vital to develop alternative ways to re-establish communication with these patients during periods of awareness, and a possible platform is through brain–computer interface (BCI).
Since consciousness is required to use BCI correctly, this study proposes a modus operandi to analyze not only in intracranial electrocorticography (ECoG) signals with greater signal-to-noise ratio (SNR) and higher signal amplitude, but also in non-invasive electroencephalography (EEG) signals. By applying three different time-domain analysis approaches sample entropy, permutation entropy, and Poincaré plot as feature extraction to prevent disease-related reductions of brainwave frequency bands in CLIS patients, and cross-validated to improve the probability of correctly detecting the conscious states of CLIS patients. Due to the lack a of 'ground truth' that could be used as teaching input to correct the outcomes, k-Means and DBSCAN these unsupervised learning methods were used to reveal the presence of different levels of consciousness for individual participation in the experiment first in locked-in state (LIS) patients with ALSFRS-R score of 0.
The results of these different methods converge on the specific periods of consciousness of CLIS/LIS patients, coinciding with the period during which CLIS/LIS patients recorded communication with an experimenter. To determine methodological feasibility, the methods were also applied to patients with disorders of consciousness (DOC). The results indicate that the use of sample entropy might be helpful to detect awareness not only in CLIS/LIS patients but also in minimally conscious state (MCS)/unresponsive wakefulness syndrome (UWS) patients, and showed good resolution for both ECoG signals up to 24 hours a day and EEG signals focused on one or two hours at the time of the experiment. This thesis focus on consistent results across multiple channels to avoid compensatory effects of brain injury.
Unlike most techniques designed to help clinicians diagnose and understand patients' long-term disease progression or distinguish between different disease types on the clinical scales of consciousness. The aim of this investigation is to develop a reliable brain-computer interface-based communication aid eventually to provide family members with a method for short-term communication with CLIS patients in daily life, and at the same time, this will keep patients' brains active to increase patients' willingness to live and improve their quality of life (QOL).
| Identifer | oai:union.ndltd.org:DRESDEN/oai:qucosa:de:qucosa:79782 |
| Date | 30 June 2022 |
| Creators | Wu, Shang-Ju |
| Contributors | Universität Leipzig |
| Source Sets | Hochschulschriftenserver (HSSS) der SLUB Dresden |
| Language | English |
| Detected Language | English |
| Type | info:eu-repo/semantics/acceptedVersion, doc-type:doctoralThesis, info:eu-repo/semantics/doctoralThesis, doc-type:Text |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0028 seconds