Rozprawy doktorskie na temat „Vision prosthesis”

Microelectronic vision prostheses aim to restore visual percepts through electrical stimulation of the surviving visual pathways in the blind. Electrical stimulation has been shown to produce spots of light in the visual field. A neurostimulator that forms the basis of a vision prosthesis was designed using a high voltage CMOS process to allow it to be able to stimulate when faced with high electrode-tissue impedances. It was implemented with novel features that allow it to be scalable, and to focus charge injection, and can stimulate multiple sites simultaneously using a current source and sink at each site. To reduce electrical cross-talk between multiple stimulation sites, six-return electrodes surround each stimulating electrode, electrically guarding them from each other. The six-return electrode configuration was shown to reduce electrical cross-talk in saline bath tests compared to single-return electrode configurations. The neurostimulator was used to evoke responses from cats through electrical stimulation via intravitreal ball electrodes, corneal electrodes, and planar electrode arrays in the suprachoroidal space. Responses were measured on the visual cortex through optical imaging of intrinsic signals, and through surface electrodes. Using the planar electrode array in the suprachoroidal space, responses were elicited to biphasic, bipolar and monopolar stimuli, with each stimulating electrode coupled with either six-return electrodes, two-return electrodes, or a single-return electrode. The average charge threshold to elicit a response for biphasic, bipolar stimulation with six-return electrodes was 76.47 ?? 8.76 nC (standard error of the mean). For biphasic, bipolar stimulation, the magnitude and area of cortical response with the six-return electrode configurations was on average 2.18 ?? 0.19 times smaller than single-return electrode configurations, and 1.89 ?? 0.19 times smaller than two-return electrode configurations (P < 0.0001). It was also found that for biphasic stimulation, a greater magnitude and area of response was elicited for monopolar stimulation compared to bipolar stimulation. This dissertation details the design and testing of a novel, scalable neurostimulator to focus charge injection. It also shows that suprachoroidal, bipolar stimulation can elicit visual responses, and that the area of cortical activation was more focused when using bipolar, biphasic stimulation, and six-return electrodes.

Restoring vision to the blind has been an objective of several research teams for a number of years. It is known that spots of light -- phosphenes -- can be elicited by way of electrical stimulation of surviving retinal neurons. Beyond this, however, our understanding of prosthetic vision remains rudimentary. To advance the realisation of a clinically viable prosthesis for the blind, a versatile integrated circuit neurostimulator was designed, manufactured, and verified. The neurostimulator provides electrical stimuli to surviving neurons in the visual pathway, affording blind patients some form of patterned vision; besides other benefits (independence), this limited vision would let patients distinguish between day and night (resetting their circadian rhythm). This thesis presents the development of the neurostimulator, an interdisciplinary work bridging engineering and medicine. Features of the neurostimulator include: high-voltage CMOS transistors in key circuits, to prevent voltage compliance issues due to an unknown or changing combined tissue and electrode/tissue interface impedance; simultaneous stimulation using current sources and sinks, with return electrodes configured to provide maximum charge containment at each stimulation site; stimuli delivered to a two dimensional mosaic of hexagonally packed electrodes, multiplexing current sources and sinks to allow each electrode in the whole mosaic to become a stimulation site; electrode shorting to remove excess charge accumulated during each stimulation phase. Detailed electrical testing and characterisation verified that the neurostimulator performed as specified, and comparable to, or better than, other vision prostheses neurostimulators. In addition, results from several animal experiments verified that the neurostimulator can elicit electrically evoked visual responses. The features of the neurostimulator enable research into how simultaneous electrical stimulation affects the visual neural pathways; those research results could impact other neural prosthetics research and devices.

Background. A visual prosthesis is a conceptual device designed to activate residual functional neurons in the visual pathway of blind individuals to produce artificial vision. Such device, when applied to stimulate the vitreous surface of the retina, has proven feasible in producing patterned light perception in blind individuals suffering from dystrophic diseases of the retina, such as aged-related macular degeneration (AMD). However the practicality of such approach has been challenged by the difficulty of surgical access and the risks of damaging the neuroretina. Positioning a visual implant over the scleral surface of the eye could present a safer alternative but this stimulation modality has not been tested in diseased retinas. Additionally, recent research has shown that the adult neocortex retains substantial plasticity following a disruption to its visual input and the potential deterioration in visual capabilities as a result of such experience modification may undermine the overall bionic rescue strategy. Methods. Two animal models mimicking the principal pathologies found in AMD, namely photoreceptor degeneration and reduced retinal ganglion cell mass, were used to evaluate the efficacy of trans-scleral stimulation of the retina by recording electrical evoked potentials in the visual cortex. The visual performance following the loss of pattern vision induced by bilateral eyelid suturing in adult mice was examined by analysing visual evoked potentials. Findings. Spatially differentiated cortical activations were obtained notwithstanding the underlying retinopathy in the experiment animals. The charge density thresholds were found to be similar to controls and below the bioelectric safety limit. After prolonged visual deprivation (weeks) in the mouse, the visual cortical responses evoked by either electrical or photic stimuli were both significantly reduced. An assessment of different visual capabilities using patterned stimuli demonstrated that whilst visual acuity and motion sensitivity were preserved, significant depression in luminance and contrast sensitivities was detected. Conclusion. Trans-scleral stimulation of the retina is a feasible approach for the development of a visual prosthesis. Following visual loss the adult brain exhibits significant experience-dependent modifications. These new insights may force a revision on the current bionic rescue strategy.

Ces travaux s'inscrivent dans le cadre plus général de la réalisation d'une prothèse visuelle, destinée aux personnes atteintes de cécité tardive due à une maladie dégénérative des cellules photo réceptrices de la rétine. Par la stimulation adéquate de certaines parties des voies optiques encore fonctionnelles, il est possible de transmettre à nouveau des perceptions visuelles, appelées phosphènes, dans le champ visuel du patient. Les études actuelles portant sur le contenu informationnel de cette vision prothétique proposent simplement de réduire la résolution de l'image d'une caméra embarquée. Notre étude propose une nouvelle approche basée sur le principe de la vision fonctionnelle, développée en collaboration avec les spécialistes de la rééducation fonctionnelle de l'institut ARAMAV. Par le biais de cette approche innovante, nous avons défini des méthodes d'extraction et de représentation des informations de la scène. L'objectif est d'améliorer l'autonomie en mobilité et de rendre possible la perception des expressions faciales. Pour évaluer la pertinence de ces propositions, nous avons développé un simulateur de vision prothétique, dont les paramètres sont basés sur les résultats d'essais cliniques. Pour la mobilité, nous mettons notamment en évidence l'intérêt de fournir un accès aux informations 3D, en substitution ou en complément de l'information de luminosité, ainsi que l'intérêt d'utiliser un oculomètre pour améliorer et faciliter la prise d'informations. Des expériences spécifiques à la mobilité, ainsi que certains résultats initiaux sont présentés. Pour la reconnaissance des expressions faciales, notre étude apporte la preuve que ces expressions sont correctement perçues si nous combinons des techniques de sélection de l'information et l'utilisation de certains prétraitements
This work falls within the broader framework of visual prostheses conception, designed for people suffering from late blindness due to degenerative diseases of retina photoreceptor cells. By stimulating certain part of the optical pathway that is still functional, it is possible to elicit visual perceptions, called phosphenes, in the subject's visual field. Recent studies on the problematic of informational content of prosthetic vision propose, in majority, a simple reduction in the resolution of grayscale images acquired from a single head worn camera. Our study proposes a new approach based on the principle of functional vision, developed in collaboration with specialists in functional rehabilitation of the ARAMAV institute. Through this innovative approach, we have defined methods for extraction and representation of scene informational content, which aim to improve autonomy in mobility and to make possible the perception of facial expressio ns. In order to evaluate these propositions, we have developed a prosthetic vision simulator, whose parameters are based on clinical trials results. For mobility, we present the usefulness of providing access to 3D information, in substitution or in addition with brightness information and the usefulness of an eye tracking device to improve and facilitate spatial knowledge acquisition. Experiments dedicated to mobility as well as some preliminary results are presented. For the recognition of facial expressions, our study provides evidence that these expressions are correctly perceived if we combined information selection and specific image processing techniques

Electrical stimulation of the retina affected by photoreceptor loss (e.g., cases of retinitis pigmentosa) elicits the perception of luminous spots (so-called phosphenes) in the visual field. This phenomenon, attributed to the relatively high survival rates of neurons comprising the retina's inner layer, serves as the cornerstone of efforts to provide a microelectronic retinal prosthesis -- a device analogous to the cochlear implant. This thesis concerns phosphenes -- their elicitation and modulation, and, in turn, image analysis for use in a prosthesis. This thesis begins with a comparative review of visual modelling of electrical epiretinal stimulation and analogous acoustic modelling of electrical cochlear stimulation. The latter models involve coloured noise played to normal listeners so as to investigate speech processing and electrode design for use in cochlear implants. Subsequently, four experiments (three psychophysical and one numerical), and two statistical analyses, are presented. Intrinsic signal optical imaging in cerebral cortex is canvassed appendically. The first experiment describes a visual tracking task administered to 20 normal observers afforded simulated prosthetic vision. Fixation, saccade, and smooth pursuit, and the effect of practice, were assessed. Further, an image analysis scheme is demonstrated that, compared to existing approaches, assisted fixation and pursuit (but not saccade) accuracy (35.8% and 6.8%, respectively), and required less phosphene array scanning. Subsequently, (numerical) information-theoretic reasoning is provided for the scheme's superiority. This reasoning was then employed to further optimise the scheme (resulting in a filter comprising overlapping Gaussian kernels), and may be readily extended to arbitrary arrangements of many phosphenes. A face recognition study, wherein stimuli comprised either size- or intensity-modulated phosphenes, is then presented. The study involved unpracticed observers (n=85), and showed no 'size' --versus--'intensity' effect. Overall, a 400-phosphene (100-phosphene) image afforded subjects 89.0% (64.0%) correct recognition (two-interval forced-choice paradigm) when five seconds' scanning was allowed. Performance fell (64.5%) when the 400-phosphene image was stabilised on the retina and presented briefly. Scanning was similar in 400- and 100-phosphene tasks. The final chapter presents the statistical effects of sampling and rendering jitter on the phosphene image. These results may generalise to low-resolution imaging systems involving loosely packed pixels.

The electrical stimulation of appropriate components of the human visual system can result in the perception of blobs of light (or phosphenes) in totally blind patients. By stimulating an array of closely aligned electrodes it is possible for a patient to perceive very low-resolution images from spatially aligned phosphenes. Using this approach, a number of international research groups are working toward developing multiple electrode systems (called Artificial Human Vision (AHV) systems or visual prostheses) to provide a phosphene-based substitute for normal human vision. Despite the great promise, there are currently a number of constraints with current AHV systems. These include limitations in the number of electrodes which can be implanted and the perceived spatial layout and display frequency of phosphenes. Therefore the development of computer vision techniques that can maximise the visualisation value of the limited number of phosphenes would be useful in compensating for these constraints. The lack of an objective method for comparing different AHV system displays, in addition to comparing AHV systems and other blind mobility aids (such as the long cane), has been a significant problem for AHV researchers. Finally, AHV research in Australia and many other countries relies strongly on theoretical models and animal experimentation due to the difficult of prototype human trials. Because of this constraint the experiments conducted in this thesis were limited to simulated AHV devices with normally sighted research participants and the true impact on blind people can only be regarded as approximated. In light of these constraints, this thesis has two general aims. The first aim is to investigate, evaluate and develop effective techniques for mobility assessment which will allow the objective comparison of different AHV system phosphene presentation methods. The second aim is to develop a useful display framework to guide the development of AHV information presentation, and use this framework to guide the development of an AHV simulation device. The first research contribution resulting from this work is a conceptual framework based on literature reviews of blind and low vision mobility, AHV technology, and computer vision. This framework incorporates a comprehensive number of factors which affect the effectiveness of information presentation in an AHV system. Experiments reported in this thesis have investigated a number of these factors using simulated AHV with human participants. It has been found that higher spatial resolution is associated with accurate walking (reduced veering), whereas higher display rate is associated with faster walking speeds. In this way it has been demonstrated that the conceptual framework supports and guides the development of an adaptive AHV system, with the dynamic adjustment of display properties in real-time. The second research contribution addresses mobility assessment which has been identified as an important issue in the AHV literature. This thesis presents the adaptation of a mobility assessment method from the blind and low vision literature to measure simulated AHV mobility performance using real-time computer based analysis. This method of mobility assessment (based on parameters for walking speed, obstacle contacts and veering) is demonstrated experimentally in two different indoor mobility courses. These experiments involved sixty-five participants wearing a head-mounted simulation device. The final research contribution in this thesis is the development and evaluation of an original real-time looming obstacle detector, based on coarse optical flow, and implemented on a Windows PocketPC based Personal Digital Assistant (PDA) using a CF card camera. PDA based processors are a preferred main processing platform for AHV systems due to their small size, light weight and ease of software development. However, PDA devices are currently constrained by restricted random access memory, lack of a floating point unit and slow internal bus speeds. Therefore any real-time software needs to maximise the use of integer calculations and minimise memory usage. This contribution was significant as the resulting device provided a selection of experimental results and subjective opinions.

The electrical stimulation of appropriate components of the human visual system can result in the perception of blobs of light (or phosphenes) in totally blind patients. By stimulating an array of closely aligned electrodes it is possible for a patient to perceive very low-resolution images from spatially aligned phosphenes. Using this approach, a number of international research groups are working toward developing multiple electrode systems (called Artificial Human Vision (AHV) systems or visual prostheses) to provide a phosphene-based substitute for normal human vision. Despite the great promise, there are currently a number of constraints with current AHV systems. These include limitations in the number of electrodes which can be implanted and the perceived spatial layout and display frequency of phosphenes. Therefore the development of computer vision techniques that can maximise the visualisation value of the limited number of phosphenes would be useful in compensating for these constraints. The lack of an objective method for comparing different AHV system displays, in addition to comparing AHV systems and other blind mobility aids (such as the long cane), has been a significant problem for AHV researchers. Finally, AHV research in Australia and many other countries relies strongly on theoretical models and animal experimentation due to the difficult of prototype human trials. Because of this constraint the experiments conducted in this thesis were limited to simulated AHV devices with normally sighted research participants and the true impact on blind people can only be regarded as approximated. In light of these constraints, this thesis has two general aims. The first aim is to investigate, evaluate and develop effective techniques for mobility assessment which will allow the objective comparison of different AHV system phosphene presentation methods. The second aim is to develop a useful display framework to guide the development of AHV information presentation, and use this framework to guide the development of an AHV simulation device. The first research contribution resulting from this work is a conceptual framework based on literature reviews of blind and low vision mobility, AHV technology, and computer vision. This framework incorporates a comprehensive number of factors which affect the effectiveness of information presentation in an AHV system. Experiments reported in this thesis have investigated a number of these factors using simulated AHV with human participants. It has been found that higher spatial resolution is associated with accurate walking (reduced veering), whereas higher display rate is associated with faster walking speeds. In this way it has been demonstrated that the conceptual framework supports and guides the development of an adaptive AHV system, with the dynamic adjustment of display properties in real-time. The second research contribution addresses mobility assessment which has been identified as an important issue in the AHV literature. This thesis presents the adaptation of a mobility assessment method from the blind and low vision literature to measure simulated AHV mobility performance using real-time computer based analysis. This method of mobility assessment (based on parameters for walking speed, obstacle contacts and veering) is demonstrated experimentally in two different indoor mobility courses. These experiments involved sixty-five participants wearing a head-mounted simulation device. The final research contribution in this thesis is the development and evaluation of an original real-time looming obstacle detector, based on coarse optical flow, and implemented on a Windows PocketPC based Personal Digital Assistant (PDA) using a CF card camera. PDA based processors are a preferred main processing platform for AHV systems due to their small size, light weight and ease of software development. However, PDA devices are currently constrained by restricted random access memory, lack of a floating point unit and slow internal bus speeds. Therefore any real-time software needs to maximise the use of integer calculations and minimise memory usage. This contribution was significant as the resulting device provided a selection of experimental results and subjective opinions.

Thesis (Ph.D.); Submitted to the Univ. of California, Davis, CA (US); 10 Jun 2003.
Published through the Information Bridge: DOE Scientific and Technical Information. "UCRL-LR-153347" Maghribi, M. 06/10/2003. Report is also available in paper and microfiche from NTIS.

For elderly Canadians, the prevalence of vision impairment caused by degenerative retinal pathologies, such as age-related macular degeneration and retinitis pigmentosa, is at an occurrence rate of 14 percent, and on the rise. It has been shown that visual function can be restored by electrically stimulating intact retinal tissue with an array of micro-electrodes with suitable signals. Commercial retinal implants carrying such a micro-electrode array achieve this, but to date must receive power and data over copper wire cable passing through a permanent surgical incision in the eye wall (sclera). This project is defined by a collaboration with iBIONICS, who are developing retinal implants for treatment of such conditions. iBIONICS has developed the Diamond Eye retinal implant, along with several technology sub-systems to form a comprehensive and viable medical solution. Notably, the Diamond Eye system can be powered wirelessly, with no need for a permanent surgical incision. The thesis work is focused on the formulation, simulation and hardware demonstration of a powering system, mounted on glasses frame, for a retinal implant. The system includes a Micro-Electro-Mechanical System (MEMS) mirror that directs a laser beam to the implant through the pupil opening. The work presented here is built on two main components: an iterative predictor-corrector algorithm (Kalman filter) that estimates pupil coordinates from measurements provided by an image-based eye tracking algorithm; and an misalignment compensation algorithm that maps eye pupil coordinates into mirror coordinates, and compensates for misalignment caused by rigid body motions of the glasses lens mirror and the MEMS mirror with respect to the eye. Pupil tracker and misalignment compensation control performance are illustrated through simulated scenarios. The project also involves the development of a hardware prototype that is used to test algorithms and related software.

A successful restoration of vision should allow the blind to look, to see and to understand. The engineering of a microelectronic vision prosthesis has come a long way over the last forty years, but the understanding of how the restored form of vision would be interpreted and functionally applied to everyday living has made little progress until recent times. Prosthetic vision is not what most people think it would be; it is a visual scene composed of relatively large, isolated, spots of light so-called "phosphenes", very much like a magnified pictorial print. This thesis dissertation seeks to obtain a complete survey of the visual description of phosphenes from the human trial reports in the literature, simulate it, obtain a measure of the functional capacity of such visual perception, and explain the measured performance against design aspects of phosphene presentation, human perception, cognition and behaviour. Specifically, "visual acuity" (VA) was assessed on normally sighted subjects (N=15) administered with "simulated prosthetic vision". VA is a functional measure of vision highly correlated to many daily activities. Aggregating the results from the study with the other VA studies in prosthetic vision, it is shown that in general, the density of the phosphene field determines the affordable VA; however, design aspects relating to the phosphene field lattice (0.03 10gMAR with the hexagonal lattice as opposed to a square lattice) and image processing routines (0.15 10gMAR at optimised settings) can be further fine-tuned to improve VA performance. Significant performance improvement also arose from learning (0.13 10gMAR over ten visitations) and visual scanning adaptation (0.20 10gMAR with a circular scanning strategy). Performance improvements are likely related to various preferences and perceptual preferences of the human visual system. A rehabilitation program targeting the appropriate behavioural adaptation coupled with image processing routine optimised for image comprehension should provide a vision prosthesis recipient with the best functional experience to restored vision.

This thesis explores methods for enhancing digital image-like sensations which might be similar to those experienced by blind users of electronic visual prostheses. Visual prostheses, otherwise referred to as artificial vision systems or bionic eyes, may operate at ultra low image quality and information levels as opposed to more common electronic displays such as televisions, for which our expectations of image quality are much higher. The scope of the research is limited to enhancement by digital image processing: that is, by manipulating the content of images presented to the user. The work was undertaken to improve the effectiveness of visual prostheses in representing the visible world. Presently visual prosthesis development is limited to animal models in Australia and prototype human trials overseas. Consequently this thesis deals with simulated vision experiments using normally sighted viewers. The experiments involve an original application of existing image processing techniques to the field of low quality vision anticipated from visual prostheses. Resulting from this work are firstly recommendations for effective image processing methods for enhancing viewer perception when using visual prosthesis prototypes. Although limited to low quality images, recognition of some objects can still be achieved, and it is useful for a viewer to be presented with several variations of the image representing different processing methods. Scene understanding can be improved by incorporating Region-of-Interest techniques that identify salient areas within images and allow a user to zoom into that area of the image. Also there is some benefit in tailoring the image processing depending on the type of scene. Secondly the research involved the construction of a metric for basic information required for the interpretation of a visual scene at low image quality. The amount of information content within an image was quantified using inherent attributes of the image and shown to be positively correlated with the ability of the image to be recognised at low quality.

This thesis explores methods for enhancing digital image-like sensations which might be similar to those experienced by blind users of electronic visual prostheses. Visual prostheses, otherwise referred to as artificial vision systems or bionic eyes, may operate at ultra low image quality and information levels as opposed to more common electronic displays such as televisions, for which our expectations of image quality are much higher. The scope of the research is limited to enhancement by digital image processing: that is, by manipulating the content of images presented to the user. The work was undertaken to improve the effectiveness of visual prostheses in representing the visible world. Presently visual prosthesis development is limited to animal models in Australia and prototype human trials overseas. Consequently this thesis deals with simulated vision experiments using normally sighted viewers. The experiments involve an original application of existing image processing techniques to the field of low quality vision anticipated from visual prostheses. Resulting from this work are firstly recommendations for effective image processing methods for enhancing viewer perception when using visual prosthesis prototypes. Although limited to low quality images, recognition of some objects can still be achieved, and it is useful for a viewer to be presented with several variations of the image representing different processing methods. Scene understanding can be improved by incorporating Region-of-Interest techniques that identify salient areas within images and allow a user to zoom into that area of the image. Also there is some benefit in tailoring the image processing depending on the type of scene. Secondly the research involved the construction of a metric for basic information required for the interpretation of a visual scene at low image quality. The amount of information content within an image was quantified using inherent attributes of the image and shown to be positively correlated with the ability of the image to be recognised at low quality.

Un million et demi de personnes souffrent de la Rétinopathie Pigmentaire, une famille de maladies héréditaires entraînant une dégénérescence de la rétine. La maladie commence par la perte de la vision nocturne et du champ visuel périphérique et mène à une cécité totale. En raison de l'hétérogénéité des mutations génétiques responsables de la maladie, des solutions visant à compenser les symptômes de la maladie émergent. Ces prothèses rétiniennes comportent trois éléments : (i) une caméra filmant la scène devant le patient, habituellement montée sur une paire de lunettes et (ii) un dispositif de stimulation qui est capable de contrôler une partie de l'activité neuronale du patient et (iii) un processeur qui implémente la transformation entre le signal de sortie de la caméra et les commandes de stimulation. Le travail présenté dans cette thèse contribue au travail de GenSight Biologics pour développer une telle prothèse rétinienne. Le project combine deux technologies récentes, une caméra neuromorphique dans laquelle chaque pixel acquiert le signal d'une manière asynchrone, et une très haute résolution temporelle, et l'optogénétique qui permet de rendre les neurones ciblés photoexcitables. Mon travail s'étend sur l'ensemble de la chaîne de traitement du signal. Nous présentons tout d'abord un algorithme extrayant les fréquences spatiales de la vidéo à partir du flux de mesures asynchrones émises par la caméra. Ensuite, nous nous concentrons sur l'Interface Cerveau-Machine en développant un modèle de la transformation reliant le signal lumineux projeté par les lunettes et les trains de potentiels d'action déclenchés par les cellules ganglionnaires de la rétine du patient
A million and a half people suffer from Retinitis Pigmentosa, a family of inherited diseases leading to degeneration of the retina. The disease begins with the loss of night vision and peripheral visual field and leads to total blindness. Due to the heterogeneity of the genetic mutations responsible for the disease, emerging solutions aim to compensate for the symptoms of the disease rather than curing it. These retinal prostheses have three elements: (i) a camera filming the scene in front of the patient, usually mounted on a pair of glasses, (ii) a stimulation device controlling a part of the neuronal activity of the patient and (iii) a processor that implements the transformation between the output signal of the camera and the stimulation commands. The work presented in this thesis contributes to the work of GenSight Biologics to develop such a retinal prosthesis. The project combines two recent technologies, a neuromorphic camera in which each pixel acquires the signal in an asynchronous manner, and a very high temporal resolution, and optogenetics which makes the targeted neurons photoexcitable. My work spans the entire chain of signal processing. We first present an algorithm extracting the spatial frequencies of the video from the asynchronous measurement stream emitted by the camera. Next, we focus on the Brain-Machine Interface by developing a model of the transformation linking the projected light signal and the trains of action potential triggered by the patient's retinal ganglion cells

La cécité touche 39 millions de personnes dans le monde et génère de nombreuses difficultés dans la vie quotidienne. Plus précisément, les capacités de navigation (incluant orientation et mobilité) sont fortement diminuées, ce qui amène les personnes non-voyantes à limiter, voire à cesser leurs déplacements. Pour restaurer des sensations "visuelles", et par-delà, une certaine autonomie, il est possible de stimuler directement le système visuel résiduel d'une personne non-voyante à l'aide d'un implant administrant des micro-stimulations électriques. Le dispositif complet se compose d'une micro-caméra portée sur des lunettes et reliée à un ordinateur de poche, qui lui-même est connecté à l'implant. Lors des micro-stimulations, les sujets perçoivent des tâches grises, blanches ou jaunâtres appelées phosphènes. Ainsi la qualité de la vision restaurée est directement dépendante de la résolution et de la position de l'implant. Le nombre d'électrodes étant faible pour les implants en développement (moins d'une centaine), il est nécessaire de réduire drastiquement la résolution du flux vidéo pour la faire correspondre à la faible résolution de l'implant. Actuellement, l'Argus II de la société Second Sight est l'implant dont le développement est le plus avancé et sa résolution est de 60 électrodes, ce qui permet aux patients implantés de percevoir 60 phosphènes différents. Cette vision restaurée est donc très pauvre et un travail d'optimisation du signal est nécessaire pour pouvoir utiliser l'implant de manière fonctionnelle. Les sujets implantés sont impliqués dans des protocoles cliniques fermés ne permettant pas de les inclure dans d'autres expériences. Malgré cela, il est possible d'étudier les possibilités offertes par ces implants visuels en simulant la vision prothétique dans un casque de réalité virtuelle porté par des sujets voyants. Il s'agit du domaine de la vision prothétique simulée (VPS). La navigation n'a jamais été étudiée chez les patients implantés et très rarement en VPS. Il s'avère qu'avec des implants de très faible résolution, elle pose de grandes difficultés liées à la mobilité mais également des difficultés liées à l'orientation. Les travaux entrepris dans ce doctorat se concentrent sur l'étude de la navigation en VPS. Différentes théories en psychologie nous ont permis d'identifier les éléments importants pour les sujets afin qu'ils se repèrent et se construisent une représentation mentale fiable de l'environnement lors de la navigation. À partir de ces modèles, différents rendus prothétiques utilisant la vision par ordinateur ont été conçus et testés dans une tâche de navigation réalisée dans un environnement virtuel. Les expérimentations effectuées avaient pour objectif d'optimiser la perception et la compréhension de l'espace parcouru avec un implant de faible résolution. Ces évaluations reposaient sur la performance de temps des sujets pour effectuer la tâche de navigation et sur leur représentation mentale de l'environnement. Après la tâche de navigation, il leur était demandé de dessiner la carte des environnements explorés, afin d'évaluer ces représentations. Cette double évaluation a permis d'identifier les indices importants permettant de faciliter la perception et la mémorisation de la structure des environnements dans une tâche de navigation en VPS. Pour améliorer les performances des personnes non-voyantes implantées, il apparaît notamment nécessaire de limiter la quantité d'information présentée, tout en préservant la structure de l'environnement grâce à des algorithmes de vision par ordinateur. Lorsque l'accès à des patients implantés sera plus ouvert, il deviendra nécessaire de valider ces différents résultats en les faisant naviguer en environnement virtuel puis en environnement réel
Blindness affects thirty nine millions people in the world and generates numerous difficulties in everyday life. Specifically, navigation abilities (which include wayfinding and mobility) are heavily diminished. This leads blind people to limit and eventually to stop walking outside. Visual neuroprosthesis are developed in order to restore such "visual" perception and help them to get some autonomy back. Those implants generate electrical micro-stimulations which are focused on the retina, the optic nerve or the visual cortex. Those stimulations elicit blurry dots called "phosphenes". Phosphenes can be mainly white, grey or yellow. The whole stimulation device contains a wearable camera, a small computer and the implant which is connected to the computer. The implant resolution and position impact directly the quality of the restored visual perception. Current implants include less than a hundred electrodes so it is mandatory to reduce the resolution of the visual stream to match the implant resolution. For instance, the already commercialized Argus II implant from the company Second Sight (Seymar, California) is the leading visual implant worldwide and uses only sixty electrodes. This means that Argus II blind owners can perceive only sixty phosphenes simultaneously. Therefore this restored vision is quite poor and signal optimization is required to get to a functional implant usage. Blind people with implants are involved in restricted clinical trials and are difficult to reach. Yet, studying those implant possibilities is at our reach by simulating prosthetic vision and displaying it in a head mounted display for sighted subjects. This is the field of simulated prosthetic vision (SPV). Navigation was never studied with people with implant, and only a few studies approached this topic in SPV. In this thesis, we focused on the study of navigation in SPV. Computer vision allowed us to select which of the scene elements to display in order to help subjects to navigate and build a spatial representation of the environment. We used psychological models of navigation to conceive and evaluate SPV renderings. Subjects had to find their way and collect elements in a navigation task in SPV inspired by video games for the blind. To evaluate their performance we used a performance index based on the completion time. To evaluate their mental representation, we asked them to draw the environment layout after the task for each rendering. This double evaluation lead us to spot which elements can and should be displayed in low resolution SPV in order to navigate. Specifically those results show that to be understandable in low vision, a scene must be simple and the structure of the environment should not be hidden. When blind people with implant will become available we will be able to confirm or deny those results by evaluating their navigation in virtual and real environments

Our bodies are arguably one of the most intimate things we will ever know. But the comfort of our own physical boundaries can be altered in various ways. In this analysis, we will look at how vision contributes to the sense of owning a body by analysing six abnormal conditions: the rubber hand illusion, phantom limbs, somatoparaphrenia, the body-swap illusion, out-of-body experiences, and heautoscopy. Examinations of these experimental or pathological conditions has granted a greater understanding of body-ownership. It was discovered that vision plays different modulatory roles, being more intricately involved in full-body ownership than in part-body ownership. Vision appears to be highly connected to self-location and first-person perspective, which both are contributing factors in projecting the sense of ownership to an external location. In part-body ownership, however, vision can be overruled by other senses, such as proprioception. Though it is still able to contribute to the illusion of projecting ownership and proprioceptive displacement to a rubber hand.

La rétinite pigmentaire est une maladie neurodégénérative héréditaire de la rétine entraînant la cécité. Des technologies restaurant la vision ont vu le jour comme les neuroprothèses visuelles et la thérapie optogénétique. Leur limitation est la résolution spatiale. La neuroprothèse visuelle IRIS I de Pixium Vision et la thérapie optogénétique de Gensigth Biologics permettent une stimulation à haute fréquence temporelle. Or, augmenter la résolution temporelle mène à une stimulation plus naturelle pouvant compenser la résolution spatiale limitée. Notre étude évalue l’apport de la vision restaurée de ces dispositifs et de la résolution temporelle. Des sujets sains portant des lunettes de stimulation simulant la vision restaurée ont réalisé des tâches quotidiennes à 60 Hz et 1440 Hz et différentes qualité de restauration. Les dispositifs permettent la réalisation des tâches proposées, avec plus de facilité pour les patients traités par thérapie optogénétique. Les patients pourraient donc recouvrir une autonomie dans la réalisation de tâches quotidiennes. Nous montrons que la qualité de la stimulation influence les performances des tâches nécessitant une acuité relativement bonne. La réalisation des tâches n’a pas été facilitée par l’augmentation de la résolution temporelle. Selon la littérature, cette dernière améliore la qualité de la perception. Nous avons élaboré une tâche de discrimination de direction de mouvement à trois vitesses. Dès 120 Hz, la résolution temporelle facilite la tâche à vitesse moyenne et élevée. Par conséquent, les vitesses des scènes visuelles de notre précédente étude étaient trop faibles pour que la résolution temporelle améliore la perception
Retinitis Pigmentosa is an inherited retinal degenerative disease leading to blindness. Vision restoration techniques have been developed as visual neuroprostheses and optogenetic therapy. The limitation of these devices is their spatial resolution. The visual neuroprosthesis IRIS I developed by Pixium vision and Gensight Biologics’ optogenetic therapy allow the visual information to be captured and stimuled with a high temporal resolution. Increasing the temporal resolution leads to a more natural vision, and should overcome the low spatial resolution. Our study evaluate the contribution to these techniques and the temporal resolution, towards usefull vision. Healthy subjects wearing goggles simulating vision arising from the devices were asked to perform everyday tasks at 60Hz and 1440Hz. The devices allow the tasks to be carried out, with greater ease for patients who would be treated with optogenetic therapy. Patients could then regain autonomy in performing daily tasks. We also show that the quality of stimulation influences tasks requiring relatively sharpness. We have not identified any facilitation in the accomplishment of these tasks through increased temporal resolution. According to the literature, an improvement in visual perception should accompany the increase in temporal resolution. As such, we set up a parametric study of the temporal frequency through a task of directional discrimination at three different speeds. From 120 Hz, the temporal resolution facilitates the task at medium and high speed. Based on these results, speeds of the visual scenes from our previous experiment were too low for temporal resolution to improve the perception

碩士
國立交通大學
電子工程學系 電子研究所
102
The vision quality of the visually impaired people treated by retinal prosthesis is largely affected by the vision resolution. Due to some related technical bottlenecks, how to make effective use of existing low-resolution screen becomes an important issue. This paper presents an eye-controlled visual aid system to achieve an efficient use of low-resolution vision by zooming in the user's interested objects (when staring, zoom-in; when closing eye, zoom-out). In order to better simulate the actual low-resolution vision of the patients, we have also designed a head-mounted eye-controlled display for the research purpose. In this thesis, the proposed algorithm is mainly about eye-tracking. The existing eye-tracking methods often have difficulties in correctly detecting pupil center from side eyes or occluded eyes. While we have proposed an emission of gradient orientation based algorithm to overcome the aforementioned problems. First, we use a specially designed dynamic threshold to remove the reflection of eye, then we emit a negative gradient orientation ray from each pixel. After obtaining this ray diagram, we extract the maximum cluster and analyze it to find the pupil center. The proposed algorithm exhibits many desired properties of front eyes, side eyes and occluded eyes. Furthermore, the method is invariant to severe lighting changes. Thus we have a desirable result in the zoom-in/out experiment.

No
PURPOSE: Epidemiologic research has shown that multifocal spectacle wearers (bifocal and progressive addition lenses [PALs]) are more than twice as likely to fall than are nonmultifocal spectacle wearers, with this risk further increasing when negotiating stairs. The present study investigated whether step and stair descent safety is improved by using single-vision distance lenses. METHODS: From a stationary standing position on top of a block, 20 long-term multifocal wearers stepped down (from different block heights) onto a lower level wearing bifocal, progressive addition, or single-vision distance lenses. RESULTS: Use of single-vision distance spectacles led to an increased single-limb support time, a reduced ankle and knee angle and vertical center-of-mass velocity at contact with the lower level, and a reduced ankle angular velocity and vertical center-of-mass velocity during initial landing (P < 0.03). These findings indicate that landing occurred in a more controlled manner when the subjects wore single-vision distance spectacles, rather than tending to "drop" onto the lower level as occurred when wearing bifocals or PALs. CONCLUSIONS: Use of single-vision distance spectacles led to improvements in landing control, consistent with individuals' being more certain regarding the precise height of the lower floor level. This enhanced control was attributed to having a view of the foot, step edge, and immediate floor area that was not blurred, magnified, or doubled and that did not suffer from image jump or peripheral distortions. These findings provide further evidence that use of single-vision distance lenses in everyday locomotion may be advantageous for elderly multifocal wearers who have a high risk of falling.

Low or impaired vision is a common cause of disability, with prevalence rates estimated to be between 2.7% and 5.8%. Those with low vision report reduced independence and social function, resulting in lower overall quality of life. While many causes of vision loss can be treated, there are still no therapies which can fully restore vision lost to retinitis pigmentosa or macular degeneration. Prosthetic vision systems aim to improve quality of life by restoring recipients' ability to carry out everyday visual tasks. When parts of the visual system are affected by disease or injury, visual prostheses attempt to replicate their function by communicating visual information to the user. However, the state-of-the-art is limited in terms of functional outcomes for users. The visual function of prosthetic vision recipients is considered, at best, profoundly low. Users may still require mobility aids such as guide dogs or long canes for orientation and mobility. Current devices can only stimulate a small number of locations in the visual field, with few discrete levels of stimulation intensity. The resulting visual stimuli have low resolution and limited perceivable contrast. This limits the functional outcomes for the user, such as navigation or object recognition, to high-contrast environments. Many prosthetic vision systems use an external camera, with an image processing system that produces stimuli by downsampling the camera image. Advances in computer vision techniques can thus be exploited to improve functional outcomes by generating more informative stimuli. In this thesis, our goal is to overcome functional shortcomings of prosthetic vision. To that end, we present, to our knowledge, the first application of semantic labelling to prosthetic vision. Semantic labelling allows the simultaneous detection and localisation of objects in an image, enabling an assistive device to understand a scene and present a simplified, task-specific representation to the user. We show how this can apply to a range of real-world orientation and mobility tasks. Computational cost is a signifi cant obstacle to implementing semantic labelling in a wearable system. We address this by contributing novel, fast pixel-wise semantic labelling techniques based on sparsely computed unary potentials. Our method allows us to trade-off labelling accuracy for reduced computational cost. We show how semantic labelling can be applied to prosthetic vision by mapping semantic classes to stimulation intensity. We also introduce a fi rst-person dataset to show how our techniques may be applied in real-world situations to improve orientation and mobility.

This thesis presents a new approach to scene representation for prosthetic vision. Structurally salient information from the scene is conveyed through the prosthetic vision display. Given the low resolution and dynamic range of the display, this enables robust identification and reliable interpretation of key structural features that are missed when using standard appearance-based scene representations. Specifically, two different types of salient structure are investigated: salient edge structure, for depiction of scene shape to the user; and salient object structure, for emulation of biological attention deployment when viewing a scene. This thesis proposes and evaluates novel computer vision algorithms for extracting salient edge and salient object structure from RGB-D input. Extraction of salient edge structure from the scene is first investigated through low-level analysis of surface shape. Our approach is based on the observation that regions of irregular surface shape, such as the boundary between the wall and the floor, tend to be more informative of scene structure than uniformly shaped regions. We detect these surface irregularities through multi-scale analysis of iso-disparity contour orientations, providing a real time method that robustly identifies important scene structure. This approach is then extended by using a deep CNN to learn high level information for distinguishing salient edges from structural texture. A novel depth input encoding called the depth surface descriptor (DSD) is presented, which better captures scene geometry that corresponds to salient edges, improving the learned model. These methods provide robust detection of salient edge structure in the scene. The detection of salient object structure is first achieved by noting that salient objects often have contrasting shape from their surroundings. Contrasting shape in the depth image is captured through the proposed histogram of surface orientations (HOSO) feature. This feature is used to modulate depth and colour contrast in a saliency detection framework, improving the precision of saliency seed regions and through this the accuracy of the final detection. After this, a novel formulation of structural saliency is introduced based on the angular measure of local background enclosure (LBE). This formulation addresses fundamental limitations of depth contrast methods and is not reliant on foreground depth contrast in the scene. Saliency is instead measured through the degree to which a candidate patch exhibits foreground structure. The effectiveness of the proposed approach is evaluated through both standard datasets as well as user studies that measure the contribution of structure-based representations. Our methods are found to more effectively measure salient structure in the scene than existing methods. Our approach results in improved performance compared to standard methods during practical use of an implant display.

Introdução: a reabilitação oral com prótese parcial removível assume um papel muito relevante no restabelecimento das principais funções do sistema estomatognático, em função da prevalência das desdentações parciais numa população cada vez mais envelhecida. Nesta perspetiva, torna-se importante conhecer e caracterizar as principais complicações associadas ao uso de próteses parciais removíveis, por forma a contribuir para a otimização da sua resolução. Objetivos: identificar e caracterizar as complicações existentes em próteses parciais removíveis em desdentações tipo classe I e II de Kennedy, assim como o processo de resolução das mesmas. Metodologia: este estudo é do tipo observacional longitudinal retrospectivo. Foi efetuada uma pesquisa dos pacientes reabilitados com prótese parcial removível, em desdentações tipo Classe I e II de Kennedy, na Clínica Universitária da Universidade Católica Portuguesa, desde o ano de 2011. Para tal foi utilizado o programa de gestão clínica NewSoft®, assim como os registos em papel da Área Disciplinar de Prótese Removível. Foi também solicitada a comparência dos pacientes em consultas de controlo para atualização de dados e registo de eventuais alterações da sua reabilitação protética. Após caracterização das complicações e da sua resolução, realizou-se uma análise comparativa com: género, idade, ano de colocação da PPR, localização da PPR e tipo de desdentação baseado na classificação de Kennedy. Comparou-se ainda a presença/identificação de patologia sistémica relevante para o metabolismo ósseo com a necessidade de rebasamento das selas. Os dados recolhidos foram tratados e sujeitos a análise estatística recorrendo ao programa IBM SPSS Statistics, v21.0.0 (Software Statistical Package for the Social Science). Resultados: A complicação biológica mais verificada foi a doença periodontal nos dentes pilares; por outro lado, a complicação mecânica mais incidente foi a necessidade de rebasamento das selas; no que diz respeito à análise comparativa, não se verificou uma relação significativa de dependência entre nenhuma das varáveis abordadas, isto é, em parâmetros como género, idade, ano de colocação da PPR, localização da PPR e tipo de desdentação (segundo Kennedy) não foi averiguada uma relação significativa com nenhuma das complicações protéticas abordadas. Conclusões: Segundo este estudo, não existe uma relação significativa entre género, idade, ano de colocação da PPR, localização da PPR e tipo de desdentação (segundo Kennedy) e cada uma das complicações protéticas abordadas, quer biológicas, quer mecânicas. São requeridos, portanto, novos estudos, com uma maior dimensão amostral, um maior tempo de follow-up e com intervenientes reabilitados noutras clínicas médico-dentárias universitárias e/ou privadas.
Introduction: the oral rehabilitation with removable partial prosthesis plays a very important role in the restoration of the main functions of the stomatognathic system due to the prevalence of partial edentulism in an increasingly aging population. In this perspective it is important to know and to characterize the main complications associated with the use of removable partial dentures in order to contribute to the optimization of their resolution. Objectives: To identify and characterize the existing failures in removable partial prosthesis in Kennedy’s class I and II type edentulism, as well as the process of resolving them. Materials and Methods: This study was a retrospective longitudinal observational study type. It was performed a survey of rehabilitated patients with removable partial dentures in Kennedy’s class I and II type edentulism in the University Clinic of Portuguese Catholic University, since the year 2011. For that it was used the clinical management program called NewSoft®, as well as the paper records of Removable Prosthodontics Subject Area. It was also requested to attend the patients in control visits for data update and registration of any changes in its prosthetic rehabilitation. After complications characterization and their resolution it was analysed a relationship with: gender, age, year of RPD placement, RPD location and type of edentulism based on Kennedy‘s classification. It was also compared the presence / identification of systemic pathology with relevance in bone metabolism with the need for saddles relining. The collected data was processed and subjected to statistical analysis using the IBM SPSS Statistics program, v21.0.0 (Software Statistical Package for the Social Sciences). Results: The most observed biological complication was periodontal disease in abutment teeth; on the other hand, the more common mechanical complication was the need for saddles relining; into the comparative analysis, there was no significant dependence relationship between any of the variables discussed; in other words in parameters such as gender, age, year of RPD placement, RPD location and type of edentulism (based on Kennedy) it wasn’t found a significant relationship with any of prosthetic complications addressed. Conclusions: According to this study, there was no significant relationship between gender, age, year of placement of PPR, the PPR location and type of edentulous (based on Kennedy) and each of the prosthetic complications addressed, whether biological or mechanical. Therefore it is required further studies with a larger sample size, a bigger follow-up time and participants rehabilitated in others dental university clinics and/or private clinics.