Dissertations / Theses on the topic 'Electrooculography'

The eye movements of expert and novice chess players were recorded by electrooculography (EOG) technique as they attempted to find the mate in fifteen tactically active simple chess positions. In the analysis part of the data gathered from EOG recordings, the effect of the drift, which is an important problem for EOG, was reduced. The processed data were converted into coordinates of the display on which the chess positions were demonstrated. In the test phase, the players were asked to fixate the white king which appeared on a different square on an empty chess board for one second. It was predicted on which square the subject fixated by the method developed. The predictions and the actual location of the white king were compared and the results revealed that EOG technique can be used reliably to track the eye movements of the chess players while they fixated on a chess board. It has been revealed that experts produced more fixations on the relevant squares than did the novices as the fixations were investigated. The difference between fixation duration was not significant across skill groups. The results provide evidence for an early perceptual encoding advantage for expert players attributable to chess experience.

This master’s thesis deals with muscles physiology of eye movement and questions of surface electromyography measurement. Physiology and analysis of eye movement are described here. Furthermore the paper deals with a variety of electrooculographic measuring techniques and also with resulting evaluation of measured signal.

The master´s thesis deals with fatigue evaluation in electrooculography records (EOG). The theoretical part focuses on electrooculography itself, fatigue and methods used for fatigue detection from EOG. The practical part includes a plan of optimal methodology for fatigue evaluation using EOG. The EOG signals are recorded during the volunteers are watching prepared scenes. Those scenes are desribed here. There is also definition of signal processing methods with relevant block diagrams. Laboratory protocol describing EOG signals recording using Biopac data acquisition system is included. Ten volunteers were measured according to this protocol and the signal database was created. In brief questionnaire volunteers were supposed to evaluate fatigue and discomfort of the measurment. Recorded signals were processed and acquired parameters were statistically evaluated. Then the parameters were discussed in terms of fatigue detection ability. On the basis of that software application was created. This application detects fatigue in selected signal. The thesis includes detailed laboratory manual for students.

Thesis (M. S.)--Psychology, Georgia Institute of Technology, 2009.
Committee Chair: Dr. Paul Corballis, Ph.D.; Committee Member: Dr. Daniel Spieler, Ph.D.; Committee Member: Dr. Eric Schumacher, Ph.D.; Committee Member: Dr. Krish Sathian, M.D., Ph.D.; Committee Member: Dr. Randall Engle, Ph.D.

Drowsiness is one of the main contributors to road crashes. This research program examines the utility of drowsiness detection based on singular and hybrid approaches using physiological signals of EEG, EOG, and ECG. Four supervised machine learning models were developed to detect drowsiness levels, using physiological features known to be associated with drowsiness and performance impairment. The ground truth was subjective sleepiness responses while performing a repetitive reaction time task. The outcome of the study indicates that the selected features provided higher performance in the hybrid approaches than the singular approaches, which could be useful for future research implications.

The electro-oculogram (EOG) is the most widely used technique for recording eye movements in clinical settings. It is inexpensive, practical, and non-invasive. Use of EOG is usually restricted to horizontal recordings as vertical EOG contains eyelid artefact (Oster & Stern, 1980) and blinks. The ability to analyse two dimensional (2D) eye movements may provide additional diagnostic information on pathologies, and further insights into the nature of brain functioning. Simultaneous recording of both horizontal and vertical EOG also introduces other difficulties into calibration of the eye movements, such as different gains in the two signals, and misalignment of electrodes producing crosstalk. These transformations of the signals create problems in relating the two dimensional EOG to actual rotations of the eyes. The application of an artificial neural network (ANN) that could map 2D recordings into 2D eye positions would overcome this problem and improve the utility of EOG. To determine whether ANNs are capable of correctly calibrating the saccadic eye movement data from 2D EOG (i.e. performing the necessary inverse transformation), the ANNs were first tested on data generated from mathematical models of saccadic eye movements. Multi-layer perceptrons (MLPs) with non-linear activation functions and trained with back propagation proved to be capable of calibrating simulated EOG data to a mean accuracy of 0.33° of visual angle (SE = 0.01). Linear perceptrons (LPs) were only nearly half as accurate. For five subjects performing a saccadic eye movement task in the upper right quadrant of the visual field, the mean accuracy provided by the MLPs was 1.07° of visual angle (SE = 0.01) for EOG data, and 0.95° of visual angle (SE = 0.03) for infrared limbus reflection (IRIS®) data. MLPs enabled calibration of 2D saccadic EOG to an accuracy not significantly different to that obtained with the infrared limbus tracking data.

The electro-oculogram (EOG) is the most widely used technique for recording eye movements in clinical settings. It is inexpensive, practical, and non-invasive. Use of EOG is usually restricted to horizontal recordings as vertical EOG contains eyelid artefact (Oster & Stern, 1980) and blinks. The ability to analyse two dimensional (2D) eye movements may provide additional diagnostic information on pathologies, and further insights into the nature of brain functioning. Simultaneous recording of both horizontal and vertical EOG also introduces other difficulties into calibration of the eye movements, such as different gains in the two signals, and misalignment of electrodes producing crosstalk. These transformations of the signals create problems in relating the two dimensional EOG to actual rotations of the eyes. The application of an artificial neural network (ANN) that could map 2D recordings into 2D eye positions would overcome this problem and improve the utility of EOG. To determine whether ANNs are capable of correctly calibrating the saccadic eye movement data from 2D EOG (i.e. performing the necessary inverse transformation), the ANNs were first tested on data generated from mathematical models of saccadic eye movements. Multi-layer perceptrons (MLPs) with non-linear activation functions and trained with back propagation proved to be capable of calibrating simulated EOG data to a mean accuracy of 0.33° of visual angle (SE = 0.01). Linear perceptrons (LPs) were only nearly half as accurate. For five subjects performing a saccadic eye movement task in the upper right quadrant of the visual field, the mean accuracy provided by the MLPs was 1.07° of visual angle (SE = 0.01) for EOG data, and 0.95° of visual angle (SE = 0.03) for infrared limbus reflection (IRIS®) data. MLPs enabled calibration of 2D saccadic EOG to an accuracy not significantly different to that obtained with the infrared limbus tracking data.
Thesis (PhD Doctorate)
Doctor of Philosophy (PhD)
School of Applied Psychology
Griffith Health
Full Text

У тези је предложена метода мерења ЕОГ сигнала на интервалу сапреклапањем временских прозора, која представља модификацијудигиталне стохастичке методе мерења (ДСММ). Хардвер којим сепредложена метода може имплементирати заснован је на рачунарскојплатформи и микроконтролерској структури. Метода је испитана бројнимсимулацијама и експериментима. Хипотеза ове тезе – да модел мерењаЕОГ сигнала на интервалу са преклапањем временских прозораомогућава пројектовање ЕОГ мерног уређаја који је отпорнији на утицајшума него ЕОГ мерни уређаји засновани на класичном дигиталноммерењу – потпуно је потврђена.
U tezi je predložena metoda merenja EOG signala na intervalu sapreklapanjem vremenskih prozora, koja predstavlja modifikacijudigitalne stohastičke metode merenja (DSMM). Hardver kojim sepredložena metoda može implementirati zasnovan je na računarskojplatformi i mikrokontrolerskoj strukturi. Metoda je ispitana brojnimsimulacijama i eksperimentima. Hipoteza ove teze – da model merenjaEOG signala na intervalu sa preklapanjem vremenskih prozoraomogućava projektovanje EOG mernog uređaja koji je otporniji na uticajšuma nego EOG merni uređaji zasnovani na klasičnom digitalnommerenju – potpuno je potvrđena.
EOG signal measurement method over an interval using overlapping timewindows, which represents the modification of the digital stochasticmeasurement method (DSMM), has been proposed in this thesis. Thehardware that can implement the proposed method is based on a computerplatform and a microcontroller structure. The method has been tested withnumerous simulations and experiments. Hypothesis of this thesis - the modelof EOG signal measurement over an interval using overlapping time windowsallows a design of EOG measuring device more resistant to the noise impactthan the EOG measuring devices based on classical digital measurement -has been fully confirmed.

Le système d'enregistrement et de lecture Medilog 9000 permet d'étudier le sommeil à domicile. Ce travail a pour but d'effectuer une analyse automatique des signaux à la sortie de l'appareil. Le traitement du signal est effectué en deux étapes: d'abord extraire les paramètres les plus informationnels en utilisant une technique de reconnaissance de formes et, plus particulièrement, une reconnaissance syntaxique - puis la quantification automatique des stades de sommeil en utilisant la programmation par arbre de décision

Ce mémoire concerne la mise en place d'un dispositif d'assistance technique aux handicapés moteurs pour l'accès aux environnements virtuels. Notre démarche consiste à apporter une adaptation logicielle à un organe de commande oculaire pour le rendre utilisable pour une tâche de navigation dans un monde virtuel. La commande oculaire Cyclope utilisée est basée sur la technologie électrooculographique. Elle permet de mesurer les mouvements oculaires et une interface logicielle composée de menus de navigation permet de transformer ces mouvements 2D en déplacements 3D. Le protocole mis en place permet d'étudier l'utilisation de Cyclope pour la navigation 3D selon des aspects objectifs comme le taux d'erreurs et selon un aspect subjectif concernant la fatigue ou la facilité d'utilisation. Des expérimentations menées avec des sujets valides et un sujet handicapé ont permis la validation de l'utilisation de Cyclope en 3D et l'identification des principales origines des erreurs de commande
This thesis presents the implementation of an assistance device dedicated to physically handicapped in order to provide access to virtual environments. Our method consists on software adaptation of an ocular control device to make it usable for a 3D navigation task in a virtual world. This ocular control device called Cyclope uses electrooculographic technology. It measures horizontal and vertical eye movements. A graphical user interface containing three navigation menus is used to transform 2D eye movements into 3D rotations and translations. An experimental protocol is used to study the use of the ocular control device for 3D navigation following an objective aspect i. E. Error rate as well as following a subjective aspect regarding tiredness and ease of use. Experiments accomplished with non disabled subjects and one handicapped subject led to the validation of using Cyclope in a 3D navigation task and to the identification of main causes of errors encountered during experimentations

Trötthet i trafiken är ett stort problem i samhället. Det är särskilt farligt att trött framföra tunga lastbilar i trafiken eftersom dessa fordon är stora och har ofta livsavgörande roller vid inblandning i trafikolyckor. För att angripa problemet har det i denna rapport studerats kring vilket sömnstadie som är lämpligast att vakna under, i syfte att vakna pigg och alert samt vilka typer av tekniker och metoder som är lämpliga för att portabelt kunna detektera Rapid-Eye-Movement. Tidigare arbeten och studier har gjorts som påvisar att uppväckning i REM-sömn är optimalt för att känna sig alert. De valda metoderna är baserade på varianter av väletablerade tekniker som används för identifiering av sömnsteg. Elektrookulografi används för att mäta ögonrörelser med hjälp av fyra elektroder som är placerade på huvudet. Kroppsrörelser upptäcks genom en accelerometer som fästs på armen. Pulsmätningar görs och används för att räkna ut pulsvariansen under sömnen. Målet är att skapa en prototyp som ska känna av när användaren är i REM-sömn och sedan väcka användaren. Detta arbete är uppdelat i två inbyggda system som görs mellan två olika examensarbeten. Resultatet blev tre sensorer som fungerar individuellt. På grund av tidsbrist och en längre felsökning blev prototypen inte färdigställd. Innan sensorerna kan tillämpas i en produkt krävs det att ytterligare tester genomförs under monitorering av en sömnspecialist.
Tiredness in traffic is a major problem in society. It is especially dangerous to drive heavy trucks when tired because these vehicles are large and often have vital roles when involved in traffic accidents. To address the problem, this degree project has studied which sleep stage is most appropriate to wake up during, in order to wake up sharp and alert, and what types of techniques and methods are suitable for portable detection of Rapid-Eye-Movement. Previous work and studies have been done which indicates that awakening during REM sleep is optimal for feeling alert. The chosen methods are based on variants of well-established techniques that are used to identify sleep stages. Electrooculography is used to measure eye movements using four electrodes placed on the head. Body movements are detected by an accelerometer attached to the arm. Pulse measurements are made and used to calculate the pulse variation during sleep. The goal is to create a prototype which will know when the user is in REM sleep and then wake the user up. This work is divided into two embedded systems that are made between two different degree projects. The result was three sensors that worked individually. Due to lack of time and a longer troubleshooting, the prototype was not completed. Before the sensors can be used in a product, additional tests are required under the supervision of a sleep specialist.

Le sommeil constitue une activité indispensable à l’adaptation et à la survie des organismes supérieurs. Chez l’être humain, il aurait un rôle dans l’intégration des expériences diurnes, de manière à préparer l’individu à la période de veille consécutive. Du fait de la pertinence adaptative des émotions, il a récemment été proposé que les émotions pré-hypniques pourraient influencer le déroulement du sommeil lequel modulerait la réactivité émotionnelle au réveil. Toutefois, ce lien bi-directionnel entre émotion et sommeil demeure peu caractérisé, notamment de par sa complexité et la difficulté à concilier les diverses approches méthodologiques. Dans ce contexte, l’objectif de la présente thèse était double : d’une part, déterminer l’impact de l’induction d’états émotionnels diurnes sur les caractéristiques du sommeil ; d’autre part, évaluer l’influence de ces états sur la réactivité émotionnelle au réveil. La programmation expérimentale a conduit à la mise en place de trois types de séquences vidéos émotionnelles, négatives, neutres ou positives, de manière à induire un état émotionnel diurne chez des participants, français ou japonais. Deux expérimentations ont eu lieu : l’une, testait l’impact des séquences émotionnelles sur l’état émotionnel des participants ; l’autre, visait à déterminer l’influence d’états émotionnels pré-hypniques sur les paramètres du sommeil et sur la réactivité à des expressions émotionnelles faciales, présentées peu après le réveil consécutif. Ces expérimentations ont permis d’obtenir plusieurs résultats novateurs. En premier lieu, il a pu être montré que les séquences vidéos induisaient des états émotionnels congruents en phase pré-hypnique. En deuxième lieu, ces états émotionnels pré-hypniques modifiaient la distribution des stades du sommeil consécutif ainsi que l’activité neurovégétative associée. En troisième lieu, nous avons pu observer que l’induction émotionnelle pré-hypnique entrainait des modifications de la réactivité émotionnelle après la période de sommeil consécutive, lors de la présentation de visages exprimant des intensités émotionnelles variables. L’interprétation des principaux résultats conduit à supposer que l’augmentation du temps passé en stade REM (Rapid Eye Movement) notamment, suite aux stimulations pré-hypniques, pourrait être l’indice d’un traitement particulier des expériences émotionnelles diurnes. Prenant appui sur cette hypothèse, il n’est pas exclu que ce traitement puisse constituer un médiateur des modifications de la réactivité émotionnelle observées lors du réveil.En conclusion, les recherches présentées dans cette thèse, fondées sur une méthodologie d’induction émotionnelle, nouvelle et standardisée, apportent des informations originales sur les liens bi-directionnels entre les émotions de la période de veille et le déroulement du sommeil. Sous réserve d’expérimentations supplémentaires, ces travaux suggèrent des questionnements nouveaux sur le rôle des émotions, notamment positives, sur le sommeil et les états émotionnels consécutifs. D’un point de vue prospectif, l’idée de l’utilisation du sommeil comme un levier efficace dans la régulation des émotions se trouve renforcée
Sleep constitutes a major component of adaptation and survival in evolved organisms. In humans, sleep could have a role in the integration of daytime experiences, preparing by this way individuals for consecutive wake period. Due to the adaptive relevance of emotions, it has recently been proposed that pre-hypnic emotions could influence the sleep course, which could modulate emotional reactivity at awakening. However, the bi-directional link between emotion and sleep remains unclear, this being explained by the complexity of such link and by the difficulty of reconciling the diversity of methodological approaches. In this context, the aim of this thesis was twofold: first, determine the impact of induced daytime emotional states on sleep characteristics; second, assess the influence of these states on the emotional reactivity at the awakening.The experimental program led to the presentation of three types of emotional movies, negative, neutral or positive, in order to induce a diurnal emotional state, in French or Japanese participants. Two experiments were carried out: one tested the impact of the emotional movies on the individual emotional states; the other determined the influence of pre-hypnic induced emotional states on sleep parameters and responsiveness to emotional facial expressions, presented to the participant at the subsequent awakening.These experiments allowed to obtain several new results. Firstly, it was shown that emotional movies induced congruent emotional states during the pre-hypnic period. Secondly, these pre-hypnic emotional states influenced the distribution of sleep stages and associated autonomic activity during the consecutive night of sleep. Thirdly, we observed that the pre-hypnic emotional induction modulated the emotional reactivity after the sleep period, during the presentation of faces expressing emotions at different intensities. The interpretation of main results can lead to the speculation that the increase of time spent in REM (Rapid Eye Movement) stage, following the pre-hypnic emotional stimulations, can be a marker of central processes related to daytime emotional experiences. Based on this assumption, such processes could be a required mediator for emotional reactivity changes observed upon awakening.In conclusion, researches presented in this thesis, subtended by a new and standardized methodology of emotional induction, provide original information about the bi-directional links between diurnal emotions and sleep course. Subject to additional experimentations, this work brings new questions about the role of emotions, particularly positive ones, on sleep and subsequent emotional states. From a prospective point of view, reported data reinforces the idea that considers the sleep as an efficient lever to regulate emotions

碩士
國立交通大學
影像與生醫光電研究所
100
In assistive research area, human-computer interface (HCI) technology is used to help disable people by conveying their intention and thinking to the outside world. Many HCI systems based on eye movement have been proposed to assistive disable people. However, due to the complexity of algorithm and difficulty of hardware implementation, there are rare general purpose designs considering the practicality and stability in real-life. Therefore, to solve these limitations and problems, a HCI system based on electrooculography (EOG) is proposed in this study. The proposed classification algorithm provides the eye state detection including fixation, saccade and blink. Moreover, in saccade detection, this algorithm can distinguish ten kind of saccade movements (i.e., up, down, left, right, much left, much right, up-left, down-left, up-right and down-right). In addition, we development a HCI system based on eye movement classification algorithm. This system provides an eye-dialing interface that can be facilitated to improve the life of disable people. The significant results are achieved that proved the performance of the proposed classification algorithm. Moreover, the EOG-based system, which can detect ten different eye movement features, is potential to be performed in real-life applications.

Dissertação de Mestrado Integrado em Engenharia Electrotécnica e de Computadores apresentada à Faculdade de Ciências e Tecnologia da Universidade de Coimbra
O reconhecimento do estado afetivo ou emocional dos humanos, pelas máquinas, assunto que se explora já há muito tempo, pretende melhorar a interação entre máquinas/computadores e humanos, adaptando-se a diferentes situações. As expressões faciais são uma das formas que permite a inferência de estados emocionais. Existem várias técnicas de reconhecimento automático de expressões faciais, que passam por técnicas baseadas em processamento de imagem ou vídeo e por utilização de biossinais, mais especificamente sinais eletromiográficos. Nesta tese, foi investigada a utilização de biossinais, nomeadamente sinais eletromiográficos e sinais eletro-oculográficos, para inferir expressões faciais e movimentos oculares. Os músculos faciais permitem detetar expressões, tais como “felicidade”, “tristeza”, “surpresa”, “raiva”, entre outras. A informação do olhar é também relevante, tanto para a criação de novas interfaces de interação homem-máquina como para a dedução de estados emocionais e sociais. Informações como: olhar para cima, olhar para baixo, desviar o olhar são úteis para este fim. Existem várias técnicas de seguimento ocular e também de deteção de piscar de olhos, nomeadamente as que se baseiam no processamento de imagem, na oculografia de infravermelhos, na técnica da bobina escleral e nos biossinais eletro-oculográficos. A fim de mimetizar as expressões faciais efetuadas pelo utilizador, bem como movimentos oculares e movimentos da cabeça (através de um sensor de inércia) foi desenvolvido um avatar 3D. O sistema de deteção incorpora três subsistemas; o primeiro deteta entre quatro expressões faciais (neutra, felicidade, tristeza e raiva), o segundo sistema deteta entre cinco movimentos oculares (“cima”, “baixo”, “direita”, “esquerda” e “piscar de olhos”) e o movimento de “franzir”. Ainda foi desenvolvido um terceiro sistema de seguimento ocular de movimentos horizontais (eye-tracking) recorrendo também a sinais eletrooculográficos. Os dois primeiros sistemas recorrem a um classificador discriminativo linear (classificador de Bayes). Todos estes sistemas foram implementados para funcionamento online. O sistema de reconhecimento de expressões faciais obteve desempenho bastante satisfatório, com uma média de exatidão, entre expressões, de 95,5%, sendo a expressão “tristeza” algumas vezes confundida pelo classificador com a expressão ”raiva” (algo que se refletiu no desempenho global do sistema).Nos testes efetuados, o sistema de reconhecimento de movimentos oculares obteve uma taxa de sensibilidade de 84,7% e um taxa de exatidão de 98,3%. O sistema de seguimento ocular, que foi desenvolvido, apresentou um ótimo desempenho, no seguimento de treze sacadas, e um erro de aproximadamente 3,0 graus na posição final.
The recognition of the human affective and emotional state by machines is a subject that has been researched for a long time. This allows for better and more suitable interaction between machines/computers and humans. Facial expressions are one of the ways that allows the inference of emotional states. Various techniques are available to automate facial expressions recognition, ranging from techniques based on the image and video processing to the use of biosignals, more precisely electromyographic signals. In this thesis, the use of biosignals, in particular electromyographic signals and electrooculographic signals, was researched to deduce facial expressions and eye movements. The facial muscles allow the detection of expressions such as “happiness”, “sadness”, “surprise”, “anger”, among other expressions. Eye information is also relevant, not only for the creation of new humanmachine interfaces, but also for depicting social and emotional states. Look in up, look in down or look away are useful information for this purpose. There are several techniques for eye tracking and detecting blinks of the eye: based on image processing, infrared oculography, scleral coil and based on electrooculographic biosignals. In order to mimic the user facial expressions, as well as eye tracking and head movements (through a motion sensor), a 3D avatar was developed. The main detecting system incorporates three subsystems. The first subsystem detects four facial expressions (neutral, happiness, sadness and anger), the second subsystem detects five eye movements (up, down, right, left and blink of the eye) and frown movement. The third subsystem is an eye-tracking system that also uses electrooculographic signals. The first two subsystems use a linear discriminative classifier (Bayes classifier). All these subsystems were developed for online usage. The facial expressions recognition system had a positive performance, with an average accuracy of 95,5%, where the sad expression was sometimes confused with anger expression (that affected the system global performance). In the tests carried out, the eye-movement recognition system had a sensitivity of 84,7% and accuracy of 98,3%. The developed eye-tracking system showed great performance when following 13 saccades, and a final position error of approximately 3,0 degrees.