The main objective of this work was the development of a computer-based Expert Sleep Analysis Methodology (ESAM) to aid sleep care physicians in the diagnosis of pre-Parkinson's disease symptoms using polysomnogram data. ESAM is significant because it streamlines the analysis of the human sleep cycles and aids the physician in the identification, treatment, and prediction of sleep disorders. In this work four aspects of computer-based human sleep analysis were investigated: polysomnogram interpretation, pre-processing, sleep event classification, and abnormal sleep detection. A review of previous developments in these four areas is provided along with their relationship to the establishment of ESAM. Polysomnogram interpretation focuses on the ambiguities found in human polysomnogram analysis when using the rule based 1968 sleep staging manual edited by Rechtschaffen and Kales (R&K). ESAM is presented as an alternative to the R&K approach in human polysomnogram interpretation. The second area, pre-processing, addresses artifact processing techniques for human polysomnograms. Sleep event classification, the third area, discusses feature selection, classification, and human sleep modeling approaches. Lastly, abnormal sleep detection focuses on polysomnogram characteristics common to patients suffering from Parkinson's disease. The technical approach in this work utilized polysomnograms of control subjects and pre-Parkinsonian disease patients obtained from the Emory Clinic Sleep Disorders Center (ECSDC) as inputs into ESAM. The engineering tools employed during the development of ESAM included the Generalized Singular Value Decomposition (GSVD) algorithm, sequential forward and backward feature selection algorithms, Particle Swarm Optimization algorithm, k-Nearest Neighbor classification, and Gaussian Observation Hidden Markov Modeling (GOHMM). In this study polysomnogram data was preprocessed for artifact removal and compensation using band-pass filtering and the GSVD algorithm. Optimal features for characterization of polysomnogram data of control subjects and pre-Parkinsonian disease patients were obtained using the sequential forward and backward feature selection algorithms, Particle Swarm Optimization, and k-Nearest Neighbor classification. ESAM output included GOHMMs constructed for both control subjects and pre-Parkinsonian disease patients. Furthermore, performance evaluation techniques were implemented to make conclusions regarding the constructed GOHMM's reflection of the underlying nature of the human sleep cycle.

Die langfristigen Folgen von psychischer Fehlbeanspruchung stellen ein beträchtliches Problem unserer modernen Gesellschaft dar. Zur Identifizierung derartiger Fehlbelastungen während der Mensch-Maschine-Interaktion (MMI) kann die objektive, kontinuierliche Messung der psychischen Beanspruchung einen wesentlichen Beitrag leisten. Neueste Entwicklungen in der Sensortechnologie und der algorithmischen Methodenentwicklung auf Basis von KI liefern die Grundlagen zu ihrer messtechnischen Bestimmung. Vorarbeiten zur Entwicklung einer Methode zur neuronalen Beanspruchungsdiagnostik sind bereits erfolgt (Radüntz, 2017). Eine praxisrelevante Nutzung dieser Ergebnisse ist erfolgsversprechend, wenn die Methode mit Wearables kombiniert werden kann. Gleichzeitig sind die Evaluation und bedingungsbezogene Reliabilitätsprüfung der entwickelten Methode zur neuronalen Beanspruchungsdiagnostik in realitätsnahen Umgebungen erforderlich. Im Rahmen von experimentellen Untersuchungen der Gebrauchstauglichkeit von kommerziellen EEG-Registrierungssystemen für den mobilen Feldeinsatz wird die darauf basierende Systemauswahl für die MMI-Praxis getroffen. Die Untersuchungen zur Validierung der kontinuierlichen Methode zur Beanspruchungsdetektion erfolgt am Beispiel des Fluglotsenarbeitsplatzes beim simulierten „Arrival Management“. / The long-term negative consequences of inappropriate mental workload on employee health constitute a serious problem for a digitalized society. Continuous, objective assessment of mental workload can provide an essential contribution to the identification of such improper load. Recent improvements in sensor technology and algorithmic methods for biosignal processing are the basis for the quantitative determination of mental workload. Neuronal workload measurement has the advantage that workload registration is located directly there where human information processing takes place, namely the brain. Preliminary studies for the development of a method for neuronal workload registration by use of the electroencephalogram (EEG) have already been carried out [Rad16, Rad17]. For the field use of these findings, the mental workload assess- ment on the basis of the EEG must be evaluated and its reliability examined with respect to several conditions in realistic environments. A further essential require-ment is that the method can be combined with the innovative technologies of gel free EEG registration and wireless signal transmission. Hence, the presented papers include two investigations. Main subject of the first investigation are experimental studies on the usability of commercially-oriented EEG systems for mobile field use and system selection for the future work. Main subject of the second investigation is the evaluation of the continuous method for neuronal mental workload registration in the field. Thereby, a challenging application was used, namely the arrival management of aircraft. The simulation of the air traffic control environment allows the realisation of realistic conditions with different levels of task load. Furthermore, the work is well contextualized in a domain which is very sensible to human-factors research.

(Spatio-temporale) Blind Source Separation (BSS) eignet sich für die Verarbeitung von Multikanal-Messungen im Bereich der kontaktlosen Biosignalerfassung. Ziel der BSS ist dabei die Trennung von (z.B. kardialen) Nutzsignalen und Störsignalen typisch für die kontaktlosen Messtechniken. Das Potential der BSS kann praktisch nur ausgeschöpft werden, wenn (1) ein geeignetes BSS-Modell verwendet wird, welches der Komplexität der Multikanal-Messung gerecht wird und (2) die unbestimmte Permutation unter den BSS-Ausgangssignalen gelöst wird, d.h. das Nutzsignal praktisch automatisiert identifiziert werden kann. Die vorliegende Arbeit entwirft ein Framework, mit dessen Hilfe die Effizienz von BSS-Algorithmen im Kontext des kamera-basierten Photoplethysmogramms bewertet werden kann. Empfehlungen zur Auswahl bestimmter Algorithmen im Zusammenhang mit spezifischen Signal-Charakteristiken werden abgeleitet. Außerdem werden im Rahmen der Arbeit Konzepte für die automatisierte Kanalauswahl nach BSS im Bereich der kontaktlosen Messung des Elektrokardiogramms entwickelt und bewertet. Neuartige Algorithmen basierend auf Sparse Coding erwiesen sich dabei als besonders effizient im Vergleich zu Standard-Methoden. / (Spatio-temporal) Blind Source Separation (BSS) provides a large potential to process distorted multichannel biosignal measurements in the context of novel contact-less recording techniques for separating distortions from the cardiac signal of interest. This potential can only be practically utilized (1) if a BSS model is applied that matches the complexity of the measurement, i.e. the signal mixture and (2) if permutation indeterminacy is solved among the BSS output components, i.e the component of interest can be practically selected. The present work, first, designs a framework to assess the efficacy of BSS algorithms in the context of the camera-based photoplethysmogram (cbPPG) and characterizes multiple BSS algorithms, accordingly. Algorithm selection recommendations for certain mixture characteristics are derived. Second, the present work develops and evaluates concepts to solve permutation indeterminacy for BSS outputs of contact-less electrocardiogram (ECG) recordings. The novel approach based on sparse coding is shown to outperform the existing concepts of higher order moments and frequency-domain features.

This thesis deals with the problem with cluster analysis and biosignal classification options. The principle of cluster analysis, methods for calculating distances between objects and the standard process in the implementation of clustering are described in the first part. For biosignals processing,it is necessary to get familiar with the primary parameters of these signals in the following sections of thesis, process biosignals and methods for recording of action potentials described. Based on studying different clustering methods is presented a program with the applied method kmedoid in the next section of this thesis. The steps of this program are described in detail and in the end of thesis functionality is tested on a database of signals ÚBMI.