No description available.

This thesis investigates the feasibility of using the non-invasive biomedical signal of oxygen saturation, or SpO₂ , to diagnose a sleep disorder known as Obstructive Sleep Apnoea Hypopnoea Syndrome (OSAHS). Epidemiologically, OSAHS is the most common condition investigated by sleep clinics. In a patient suspected of having the disorder, the upper airway is obstructed during sleep and a cessation in respiration results. An apnoea is defined as a temporary cessation of breathing. Similarly, a hypopnoea is defined as any reduction in breathing (i.e., less severe than an apnoea). The work has three main objectives; the first being to establish automated evaluation procedures for methods of quantifying apnoeic activity from the SpO₂ signal, the second being to accurately identify apnoeic and normal activity on a minute-by-minute basis, the third being to create a Respiratory Disturbance Index (RDI) based on the analysis which is comparable to the gold-standard Apnoea Hypopnoea Index (AHI) derived by experts. The detection of apnoeic activity is determined using three separate analyses: time domain, frequency domain, and autoregressive modelling with an incorporated amplitude criterion. A training dataset is utilised for algorithm development, and an independent dataset is employed for testing . All three methods result in comparable overall classification accuracies of: 81.2% (time domain), 82.1% (frequency domain), and 80.0% (autoregressive modelling with amplitude). In addition, particular attention is given to the resultant sensitivity, specificity, and accuracy values partitioned according to patient category; i.e., patients with OSAHS may be divided into normal, mild, moderate and severe. Lastly, a simple RDI is computed based on the automated analyses; i.e., the number of apnoeic segments detected divided by the total number of segments used. A comparison between computed RDI and AHI values for the test database show correlation values above 0.8. In conclusion, this thesis shows that through the automated analysis of the SpO₂ signal, OSAHS severity in patients suspected of having the disorder can be quantified. The AR-modelling with an incorporated amplitude criterion, in particular, shows the most promise for further work in this area.

L’épilepsie est une pathologie chronique. Elle se définit par la répétition de manifestations cliniques paroxystiques appelées aussi crises d’épilepsie. Ces crises résultent d’un dysfonctionnement cérébral généralisé ou focalisé dû à une décharge électrique anormale. L’électroencéphalographie (EEG) est la méthode de référence permettant l’enregistrement de l’activité électrique du cerveau. Lors d’une crise d’épilepsie, les signaux EEG prennent une allure sinusoïdale et se caractérisent par une grande intensité. Génératrice de signaux EEG, les populations neuronales se synchronisent pour créer un système dynamique. Les changements d’état des oscillations des signaux EEG, variables au cours de temps, passage d’un état stable à un état instable, reflètent le début d’une crise d’épilepsie.Dans ce cadre, l’objectif de cette thèse est de proposer une nouvelle approche de prédiction de l’avènement d’une crise d’épilepsie et de localisation des générateurs corticaux associés tout en employant un minimum d’électrodes crâniennes. Cette approche permet d’alerter le malade et son entourage afin qu’ils puissent prendre les précautions nécessaires.Pour assurer une détection précoce des prémices de la crise et une localisation précise de ses origines focales éventuelles, nous proposons une méthode fiable basée en premier lieu sur la modélisation autorégressive multivariable des signaux EEG. Cette modélisation génère des coefficients capables de décrire les changements de l’état de ce système dynamique. Une Analyse en Composantes Principales basée sur l’extraction des valeurs propres du système a été utilisée pour calculer un Indice de stabilité. La variation temporelle de cet indice permet de déterminer l’état de stabilité du système avant, pendant et après la crise d’épilepsie et de détecter d’éventuelles anomalies paroxystiques précritiques. Nos principales contributions sont comme suit :- La modélisation autorégressive et l’analyse de la stabilité pour la détection précoce de la survenue des crises d’épilepsie tout en utilisant un nombre minimal d’électrodes EEG crâniennes.La méthodologie proposée comporte quatre phases principales : un prétraitement adapté pour améliorer la qualité du signal, une extraction des paramètres pertinents du modèle autorégressif, un calcul de l’Indice de stabilité et une analyse des périodes de crises.La fiabilité de notre méthode et la pertinence de nos résultats ont a été prouvée en les comparant avec d’autres méthodes rapportées par l’état de l’art et validées sur la même base de données (CHB-MIT).- Meilleur localisation spatiotemporelle des régions des décharges électriques épileptiques sur le cortex, suite à l’amélioration de la résolution surfacique de l’EEG par l'intégration des électrodes virtuellesLes décharges électriques naissent dans des points du cortex cérébral et se propage vers d’autres points du même hémisphère ou d’un autre hémisphère. Le suivi de la propagation de ces décharges permet de contrôler l’état de conscience du malade. Il s’agit de différencier une crise focale d’une crise généralisées qui se manifestent par une altération de conscience. L’étude que nous avons menée consiste à localiser de manière fiable et précise les régions cérébrales impliquées dans la crise et à suivre leurs évolutions au cours du temps. Cette étude comporte trois phases principales :Dans un troisième temps, le suivi de l’évolution spatiotemporelle des décharges épileptiques détectées par les électrodes permet d’évaluer l’état du malade et de prédire d’éventuelles altérations de la conscience. Dans les cas d’un décharge électrique épileptique du lobe fronto-temporale ou de plusieurs régions sur les deux hémisphères, le malade passe par une altération momentanée de conscience / Epilepsy is a chronic pathology defined by the repetition of clinical paroxysmal manifestations called crises. These crises are the result of a generalized or focal brain dysfunction due to an abnormal electrical discharge. The electroencephalography (EEG) is the reference method allowing the recording of the electrical activity of the brain. During an epileptic crisis, EEG signals are of sinusoidal nature and characterized by a high intensity. Generator of the EEG signals, the neuronal populations sync to create a dynamic system. The changes of the oscillations state of the EEG signals, variables over the time, passage of a stable state to an unstable state, reflect the beginning of an epileptic crisis.The conceptual framework of this thesis describes the objective to propose a new approach to predict the occurrence of epileptic crisis and localize the cortical generators associated with the minimum of cranial electrodes. This approach allows alerting the patient and his entourage so that they can take necessary precautions.To ensure early detection of the crisis onset and a precise location of its focuses, we propose a reliable method based firstly on the Multivariate Autoregressive modeling of EEG signals. This modeling generates coefficients capable to describe early changes in the dynamic system state. A Principal Components Analysis based on the extraction of values own of the system has been used to calculate an index of stability. The temporal variation of this index is used to determine the stability of system before, during and after epilepsy and to detect any paroxysmal abnormalities precritical. Our main contributions are as follows:- The Autoregressive Modeling and stability analysis for the early detection occurrence of seizures by using a minimum number of cranial EEG electrodesThe proposed methodology has four main phases: a pretreatment adapted to improve the quality of the signal, an extraction of the relevant parameters of the autoregressive model, a calculation of the stability index and the analysis of crises periods.The reliability of our method and the relevance of our results have been proved by comparing them with other methods reported by the state of the art and validated on the same database (CHB-MIT).- Best spatiotemporal localization of epileptic electric discharge regions on the cortex, following the improvement of the surface resolution of the EEG by the integration of virtual electrodesElectrical discharges are born in the points of the cerebral cortex and propagated to other points in the same hemisphere or of the other hemisphere. The monitoring of the spread of these discharges allows controlling the state of consciousness of the patient. The study that has been conducted is to locate the brain regions involved in the crisis reliably and precisely and monitoring their evolution over time. This study consists of three main phases:In a third phase, the spatiotemporal evolution of the epileptic discharges detected by the electrodes makes it possible to evaluate the patient's condition and to predict possible alterations in consciousness. In the case of an epileptic electric discharge of the fronto-temporal lobe or of several regions on the two hemispheres, the patient passes through a momentary alteration of consciousness

Early failure detection has been an integral part of condition monitoring of critical systems, such as wind turbines and helicopter rotor drivetrains. An unexplored application of early failure detection is fatigue testing of driveline components. On many occasions, driveline components fail catastrophically, leaving no evidence of the root cause of failure and causing extensive damage to test equipment. This can be prevented by detecting failure in its early stages. Test specimen would be preserved, enabling correlation of test results with design predictions. In this thesis, a method for early failure detection of gear fatigue is proposed. The gears in questions are parts of driveline components undergoing fatigue tests. The proposed method includes generation of an autoregressive model from a healthy, time synchronously averaged vibration signal. The parameters of the generated model are then used to construct a filter, which predicts deviations from the healthy signal. The output of this filter is then processed to detect failure. Vibration data from four run to failure tests were analysed. While the proposed method detected failure in all four data sets, performance was better in tests carried out at high torque and low speed in comparison to tests carried out under low torque and high speeds. Finally, potential improvements in the proposed method to increase its effectiveness are proposed. / “Early Failure Detection” (tidig detektion av utmattningsbrott) har länge varit en viktig del av tillståndsövervakning av kritiska system, som till exempel vindkraftverk och drivsystem för rotorblad på helikoptrar. Ett mindre utforskat område av “Early Failure Detection” är utmattningstestning av komponenter för transmissionssystem. Ofta går komponenterna sönder på ett sådant sätt att grundorsaken till haveriet inte går att fastställa, och som riskerar att skada testriggarna. Detta kan förebyggas om haveriet kan upptäckas i ett tidigt skede innan komponenten gar sönder helt och hållet. Testobjeket kan då bevaras, vilket ger möjligheter att korrelera testresultatet till utmattningsberäkningar av konstruktionen.  I den här uppsatsen föreslås en metod för Early Failure Detection för drevsatser i växlar. Växlarna ingår i transmissionssystem som utmattningsprovas. Den föreslagna metoden innebär att en autoreggresiv modell skapas från en tids-synkron medelvärdesbildning på den uppmätta signalen för den oförstärda komponenten. Parametrarna från den modellen används sedan för att skapa ett filter som predikterar avvikelser mot den oförstörda komponenten. Slutligen behandlas utsignalen fran det filteret för att upptäcka utmattningsskador pa drevsatsen i växeln.  Vibrationsdata fran fyra utmattningsprov har analyserats. I samtliga prov har provet körts tills brott har konstaterats. Utmattningsskador kunde konstateras tidigt, innan brottet inträffade, i tre av de fyra fallen. Slutligen föreslås förslag på utveckling av den använda metoden for att förbättra predikteringarna.

This dissertation presents algorithms for restoring some of the major corruptions observed in archived film or video material. The two principal problems of impulsive distortion (Dirt and Sparkle or Blotches) and noise degradation are considered. There is also an algorithm for suppressing the inter-line jitter common in images decoded from noisy video signals. In the case of noise reduction and Blotch removal the thesis considers image sequences to be three dimensional signals involving evolution of features in time and space. This is necessary if any process presented is to show an improvement over standard two-dimensional techniques. It is important to recognize that consideration of image sequences must involve an appreciation of the problems incurred by the motion of objects in the scene. The most obvious implication is that due to motion, useful three dimensional processing does not necessarily proceed in a direction 'orthogonal' to the image frames. Therefore, attention is given to discussing motion estimation as it is used for image sequence processing. Some discussion is given to image sequence models and the 3D Autoregressive model is investigated. A multiresolution BM scheme is used for motion estimation throughout the major part of the thesis. Impulsive noise removal in image processing has been traditionally achieved by the use of median filter structures. A new three dimensional multilevel median structure is presented in this work with the additional use of a detector which limits the distortion caused by the filters . This technique is found to be extremely effective in practice and is an alternative to the traditional global median operation. The new median filter is shown to be superior to those previously presented with respect to the ability to reject the kind of distortion found in practice. A model based technique using the 3D AR model is also developed for detecting and removing Blotches. This technique achieves better fidelity at the expense of heavier computational load. Motion compensated 3D IIR and FIR Wiener filters are investigated with respect to their ability to reject noise in an image sequence. They are compared to several algorithms previously presented which are purely temporal in nature. The filters presented are found to be effective and compare favourably to the other algorithms. The 3D filtering process is superior to the purely temporal process as expected. The algorithm that is presented for suppressing inter-line jitter uses a 2D AR model to estimate and correct the relative displacements between the lines. The output image is much more satisfactory to the observer although in a severe case some drift of image features is to be expected. A suggestion for removing this drift is presented in the conclusions. There are several remaining problems in moving video. In particular, line scratches and picture shake/roll. Line scratches cannot be detected successfully by the detectors presented and so cannot be removed efficiently. Suppressing shake and roll involves compensating the entire frame for motion and there is a need to separate global from local motion. These difficulties provide ample opportunity for further research.