Dissertationen zum Thema „Dynamic full field optical coherence tomography“

Optical Coherence Tomography (OCT) is a non-invasive volumetric imaging technique that offers micron scale resolution over several millimetres of depth penetration in tissue. The aim of the project was to develop Full Field Swept Source OCT (FFSSOCT), charting the specifications of the system throughout and drawing comparisons with available Fourier domain OCT (FDOCT) systems. A super luminescent diode (SLD) light source was tuned to different optical frequencies using an acousto-optical tunable filter. The successive optical frequencies were distributed across the sample in parallel and the sequential interference spectra were recorded with a high speed digital camera. Using a 5mW optical source at 850nm, 70dB sensitivity was ultimately achieved in a single five second acquisition, improving the original performance by 5dB and increasing acquisition and processing time significantly. Ex vivo retinal images of rat and tree shrew were recorded with multiple layers visible. All software was custom written in Labview, improving the user interface and processing time over the existing Matlab code. To enhance the performance of the system, the camera was subsequently upgraded from 15 to 45% quantum efficiency and from 250k to 1.3M pixels. The light source was also upgraded to 20mW. The software was improved with spectral processing and dispersion compensation. 85dB sensitivity was ultimately achieved. Further ex vivo retinal images were taken, showing comparative image quality to those of the same retinal samples recorded with FDOCT. Further attempts to increase the system performance were limited by internal reflections and interfering surfaces within the FFSSOCT design leading to saturation of the digital camera. In vivo imaging was attempted using software based phase jitter compensation. Improvement could be seen with simple reflecting structures.

Cette thèse a pour objectif l’étude d’un lien effectif potentiel entre la motilité cellulaire, la mécanique cellulaire, et l’activité biochimique de ces mêmes cellules. Ce couplage a été étudié dans divers systèmes biologiques, et aussi bien dans des cultures de cellules qu’à l’intérieur de tissus plus complexes. Notamment, nous avons particulièrement cherché à détecter un couplage électromécanique dans des neurones qui pourrait être impliqué dans la propagation du message nerveux.Pour ce faire, nous avons dû développer deux microscopes optiques à la sensibilité extrême. Ces microscopes se composent de deux parties principales. La première sert à détecter des mouvements axiaux plus petits que la longueur d’onde optique, soit en dessous de 100 nanomètres. La deuxième partie permet la détection d’un signal de fluorescence, offrant la possibilité de suivre l’évolution biochimique de la cellule. Avec ces deux microscopes multimodaux, il est donc possible de suivre de manière simultanée un contraste de motilité, un contraste mécanique, un contraste structurel et un contraste biochimique. Si l’un de ces systèmes est basé sur la tomographie de cohérence optique plein champ et permet de faire de telles mesures en 3-D et en profondeur dans les tissus biologiques, le second ne permet que des mesures dans des cultures de cellules, mais est bien plus robuste au bruit mécanique. Dans ce manuscrit, nous allons essentiellement décrire le développement de ces deux appareils, et préciser les contrastes auxquels ils sont sensibles spécifiquement.Nous développerons également deux des applications principales de ces microscopes que nous avons étudié dans le détail au cours de cette thèse. La première application développe l’intérêt d’un de nos microscopes pour la détection sans marquage des principaux composants cellulaires et structuraux de la cornée et de la rétine. La seconde application tend à détecter et à suivre des ondes électromécaniques dans des neurones de mammifères
This PhD project aims to explore the relationship that might exist between the dynamic motility and mechanical behavior of different biological systems and their biochemical activity. In particular,we were interested in detecting the electromechanical coupling that may happen in active neurons, and may assist in the propagation of the action potential. With this goal in mind, we have developed two highly sensitive optical microscopes that combine one modality that detects sub-wavelength axial displacements using optical phase imaging and another modality that uses a fluorescence path. Therefore, these multimodal microscopes can combine a motility, a mechanical,a structural and a biochemical contrast at the same time. One of this system is based ona multimodal combination of full-field optical coherence tomography (FF-OCT) and allows the observation of such contrast inside thick and scattering biological tissues. The other setup provides a higher displacement sensitivity, but is limited to measurements in cell cultures. In this manuscript, we mainly discuss the development of both systems and describe the various contrastst hey can reveal. Finally, we have largely used our systems to investigate diverse functions of the eye and to look for electromechanical waves in cell cultures. The thorough description of both biological applications is also provided in the manuscript

La tomographie de cohérence optique plein champ est une technique de microscopie permettant d’imager un plan d’intérêt en profondeur dans un milieu diffusant. Cette technique a été utilisée pour l’examen de pièces opératoires dans un but de diagnostic en cancérologie. L’utilisation de cette technique permettrait en effet de fournir un outil de diagnostic peropératoire rapide et fiable, évitant ainsi de nombreuses procédures de réopération. Ces réopérations peuvent survenir lorsque – lors du diagnostic final par analyse de coupes histologiques – le pathologiste décèle la présence de tissus cancéreux restant, non retirés au cours de l’opération.L’OCT plein champ a montré de bons résultats pour cette application. Néanmoins, cette technique ne fournit qu’un contraste morphologique des tissus, ne permettant pas d’utiliser des critères de qualification des pièces opératoires basées – par exemple – sur la morphologie ou la densité cellulaire.Nous avons développé une nouvelle modalité d’imagerie basée sur l’OCT plein champ permettant de révéler un contraste métabolique dans le tissu à une échelle subcellulaire. Ce contraste permet de révéler les cellules précédemment non distinguées en OCT plein champ. Nous avons également utilisé la mesure quantitative de cette modalité pour réaliser des outils d’aide au diagnostic utilisant des approches d’apprentissage par ordinateur
Full filed optical coherence tomography is a microscopy imaging technique allowing to image a specific slice in a scattering medium, in depth. This technique has been used for the diagnosis of biopsy in cancerology. This technique could be an efficient and fast way to diagnose excised tissues during surgery. This would avoid numerous reoperations procedures. These reoperations are necessary when a pathologist suspects cancerous tissue to still be present in the patient, based on histological slide examination.FFOCT has shown promising results for that purpose. Nevertheless, this technique only gives a morphological contrast of tissues, which is not enough for applying some diagnostic criteria such as cell morphology or cell density.We developed a new imaging modality based on FFOCT allowing to reveal metabolic contrast in tissues at the subcellular scale. This contrast reveals cells previously indistinguishable with FFOCT. We also used this quantitative metric to propose tools to facilitate diagnosis, using machine learning approaches

Mon projet de doctorat a été consacré à la conception et à l'application de techniques de microscopie optique non invasives, tridimensionnelles, sans marquage et en direct pour la modélisation des maladies rétiniennes à haute résolution. Aux frontières de la physique, de l'ingénierie et de la biologie, la principale question qui a motivé mon travail était de caractériser la santé de cellules dans des tissus complexes, avec une perturbation externe minimale. En particulier, j'ai contribué au développement d'une nouvelle modalité d'imagerie appelée tomographie par cohérence optique plein champ dynamique (DFFOCT), avec une attention particulière pour les échantillons rétiniens. Dans les deux premiers chapitres de mon manuscrit, je placerai cette modalité dans le contexte d'autres technologies alternatives et je décrirai le module que j'ai co-développé. Pendant mon doctorat, j'ai travaillé sur la transformation des besoins des biologistes en matière d'observation, de contrôle et d'optimisation de leurs modèles expérimentaux en solutions optiques et techniques. Les expériences et l'analyse des données ont été principalement réalisées sur des échantillons de rétine, y compris des organoïdes sains et malades, et des explants animaux. Des études longitudinales et tridimensionnelles de modélisation de la maladie in vitro ont été réalisées, en particulier sur la dystrophie rétinienne et la dégénérescence maculaire liée à l'âge (DMLA), pour lesquelles je présenterai des résultats préliminaires dans le chapitre 2. Bien que très puissant pour étudier la dynamique cellulaire dans les tissus complexes, le DFFOCT était initialement limité aux échantillons épais, ce qui empêchait son utilisation sur des cultures cellulaires en 2D. J'ai également contribué au développement d'une autre configuration optique qui permet d'imager des cellules à proximité de lamelles de verre. En utilisant un design auto-référencé, il est devenu possible d'imager des cultures cellulaires en 2D. J'ai contribué à la réalisation de la nouvelle installation et j'ai effectué une partie des expériences de preuve de concept sur des fibroblastes humains. De plus, cette modalité a été utilisée pour établir un nouveau pipeline de discrimination cellulaire en 2D, dont les premiers résultats sont présentés dans le troisième chapitre de ce manuscrit. Puisque le DFFOCT capture les mouvements intracellulaires pour quantifier l'activité cellulaire locale, nous nous sommes demandé si nous pouvions détecter des changements physiologiques par le biais d'un changement d'activité. En travaillant sur des échantillons de rétine, nous avons essayé de détecter la réponse DFFOCT à la photo-stimulation sur des organoïdes rétiniens naturels et génétiquement modifiés, ainsi que sur des explants rétiniens.Les résultats n'ont pas encore abouti à une conclusion claire, comme nous le verrons dans le chapitre 4. Cependant, le sujet est toujours étudié au sein du groupe, car d'autres membres de l'équipe tentent d'extraire des informations significatives des ensembles de données analysés. Comme dernière étape de mon projet de doctorat, j'ai construit un nouveau setup optique : un OCT à domaine spectral (SDOCT) à coupler au DFFOCT et au microscope, décrit dans le dernier chapitre de la thèse. Le but est d'ajouter une vue macroscopique perpendiculaire à basse résolution au système existant, afin d'avoir un balayage rapide de l'échantillon entier avant de passer à l'analyse à haute résolution avec le DFFOCT. Cela nous permettrait de cibler le domaine du dépistage à haut débit, étant donné que le balayage volumétrique DFFOCT, qui prend du temps, serait remplacé par une imagerie OCT hybride SD + FF. En outre, grâce à l'algorithme dynamique, il est possible de récupérer des informations métaboliques au niveau macroscopique, comme pour le FFOCT. Les applications futures du système impliquent l'automatisation du processus d'acquisition et, éventuellement, la détection de la réponse à la photo-stimulation
My PhD project was devoted to the conception and application of non-invasive, three-dimensional, label free, live optical microscopy techniques for retinal disease modelling at high resolution. At the frontiers between physics, engineering, and biology, the main question that motivated my work was how to characterize the health of single cells in complex tissues with minimal external perturbation. In particular, I contributed to the development of a new imaging modality named dynamic full field optical coherence tomography (DFFOCT) with a particular focus on retinal samples. This new modality enabled me to study the physiology of several state-of-the art biological models, including organoids, and advanced disease models. In the two first chapters of my manuscript, I will put this new modality in context of other alternative technologies, and describe a module that I co-developed.During my PhD, I worked on transforming the needs of biologists to observe, control and optimize their experimental models into optical and technical solutions. Experiments and data analysis were mainly performed on retinal samples, including healthy and diseased organoids, and animal explants. Longitudinal and three-dimensional disease modelling in vitro studies were performed, particularly on retinal dystrophy and age-related macular degeneration (AMD), on which I will present preliminary results in chapter 2. Although very powerful to study cell dynamics in complex tissues, DFFOCT was initially limited to thick specimens, which prevented its use on 2D cell cultures, preventing to make a direct link between 2D and 3D models. During my PhD, I also contributed to the development of another optical configuration that allows to image cells close to glass coverslips. By using a self-referenced design, 2D cell cultures attached to a coverslip have become possible to image. I contributed to the realization of the modified setup and carried out part of the proof-of-concept experiments on human fibroblasts. Moreover, this new modality was used to establish a new cell discrimination pipeline in 2D, with the first results shown in the third chapter of this manuscript. Since our imaging modality captures intracellular movements within cells to quantify their local activity, we asked the question whether we could detect physiological changes through a change in activity. Working on retinal samples, we tried to detect the DFFOCT response to photo-stimulation on natural and genetically modified retinal organoids, and retinal explants. The results did not lead to a clear conclusion during my PhD as will be covered in chapter 4. However, the topic is still deeply studied in the group, as other members of the team are trying to retrieve meaningful information from the analysed datasets. As the final step of my PhD project, I built a new additional part of the optical setup: a spectral domain OCT (SDOCT) to couple to the DFFOCT and to the microscope, consequently, described in the final chapter of the thesis. The aim of this association is to add a low-resolution macroscopic perpendicular view to the existing system, in order to have a quick scan of the whole sample before delving into high resolution analysis with the DFFOCT. This would allow us to target the high content screening domain, as the time consuming DFFOCT volumetric scanning would be replaced by a hybrid SD + FF OCT imaging. Moreover, with the implementation of a dynamic algorithm, it is possible to retrieve metabolic information at macroscopic level too, similarly to what is done with the FFOCT. Future applications of the aforementioned system involve automatization of the acquisition process and, possibly, detection of photo-stimulation response

Le cancer est une des principales cause de décès dans le monde et donc un problème majeur de santé publique. Plusieurs techniques d'imagerie biomédicale servent à la recherche et aux efforts cliniques pour améliorer le pronostic du patient. Nous étudions l'utilisation d'une nouvelle famille de techniques d'imagerie, la tomographie par cohérence optique plein champ statique et dynamique, qui permet une analyse du tissu plus rapide que la technique de référence en histopathologie. Afin de faciliter l'interprétation de cette nouvelle imagerie, nous développons plusieurs méthodes exploratoires basées sur des données issues d'études cliniques. Nous proposons une méthode analytique pour une meilleure caractérisation du signal interférométrique dynamique brut, ainsi que de multiples méthodes d'aide au diagnostic à partir des images. Pour cela, des réseaux neuronaux convolutifs ont été exploités sous différents paradigmes: (i) apprentissage entièrement supervisé, dont la capacité de prédiction dépasse la performance du pathologiste; (ii) apprentissage par instances multiples, qui permet de surmonter le manque d’annotations d’experts; (iii) apprentissage contrastif, qui exploite la multi-modalité des données. Nous portons une grande attention à la validation et au décryptage des modèles boîte noire pour garantir leur bonne généralisation et enfin trouver des biomarqueurs spécifiques
Cancer is a leading cause of death worldwide making it a major public health concern. Different biomedical imaging techniques accompany both research and clinical efforts towards improving patient outcome. In this work we explore the use of a new family of imaging techniques, static and dynamic full field optical coherence tomography, which allow for a faster tissue analysis than gold standard histology. In order to facilitate the interpretation of this new imaging, we develop several exploratory methods based on data curated from clinical studies. We propose an analytical method for a better characterization of the raw dynamic interferometric signal, as well as multiple diagnostic support methods for the images. Accordingly, convolutional neural networks were exploited under various paradigms: (i) fully supervised learning, whose prediction capability surpasses the pathologist performance; (ii) multiple instance learning, which accommodates the lack of expert annotations; (iii) contrastive learning, which exploits the multi-modality of the data. Moreover, we highly focus on method validation and decoding the trained "black box" models to ensure their good generalization and to ultimately find specific biomarkers

Optical coherence tomography (OCT) is a versatile and powerful imaging technique widely used in biomedical applications. It employs non-destructive radiation and performs non-contact micrometre-scale cross-sectional imaging of the sample structure. However, classic OCT systems generally apply a single-point detection scheme, which creates inefficiencies in terms of the experimental alignment of system, the point-by-point signal acquisition process and the measurement speed. One of its variants, full-field optical coherence tomography (FF-OCT) employs parallel illumination and directly acquires en-face images with a complementary camera, hence omitting the need of electromechanical lateral scans as in classic OCT systems. Current FF-OCT systems could offer more efficient measurement procedures as well as superior imaging performance, however, they are neither economically viable nor universally applicable to different applications. There is a need for a simplified low cost system to make such a powerful technology readily available for a wide range of applications. In this thesis, the development of a low cost simple FF-OCT system is described from its system setup, experimental procedures, data analysis, and system performance. The system consists of only essential components including probing lens and a beam-splitter, together with a low cost infrared LED source and CMOS camera. During the measurement, the system only requires to control the axial movement of the sample arm and the image acquisition by the camera. For the imaging of a sample with a depth of 100 um, the FF-OCT measurement only takes less than two minutes. The time-efficient measurement with the simple system offers great advantage over the developed phase-shifting FF-OCT system, which requires lengthy measurement and excessive operations, despite the decoupling of signal strength and instantaneous phase with penetration depth. Therefore, compared to state-of-the-art systems, it has the advantage of being low-cost, fast image acquisition speed and simple experimental operations. For the data analysis of tomographic imaging, the axial position of a structural feature is determined by that of the envelope, which is obtained by processing raw FF-OCT signal with Hilbert transform. The imaging performance of the simple system is measured to have a spatial resolution of 3.6 x 10.3 um2 (axial x lateral) and a system sensitivity of 74 dB. The characterisation of small-size pharmaceutical pellet coatings, bovine corneal layers and paint films is to demonstrate the potential of the simple FF-OCT system for the tomographic imaging. The layered structures and internal morphology features can be revealed by analysing the measured FF-OCT B-scan images and A-scan signals. First of all, the simple FF-OCT system is capable of performing accurate and quick measurements of pellet coatings, which are validated by the XuCT technique. FF-OCT imaging can provide a spatial characterisation of coating layers, an accurate determination of coating thickness, and an estimation of coating uniformity and porosity, making the simple system a powerful tool for the coating evaluation of similar pharmaceutical pellets. Secondly, the simple system can detect corneal surfaces and the two anterior layers of bovine cornea. This could permit the prediction of the corneal oedematous state and epithelial erosions by the analysis of the FF-OCT results of the corneal structure. Thirdly, the simple system is capable of revealing the surface and subsurface of basecoat and clearcoat films. The measurement of their paint thicknesses is also verified by the reference profilometry results. FF-OCT imaging can provide further spatial evaluation of a paint film and the areal thickness map could be obtained. The study of these paint samples with the simple system might provide an indication for the FF-OCT measurement of industrial automotive paint. For the data analysis of the surface topography, the axial position of the surface is obtained by applying interpolation and a minimum search algorithm to the raw FF-OCT signal. This allows sub-micrometre depth precision to be obtained with the simple system. In the validation of the measurement of the surface topography, a less than 10 nm deviation of the FF-OCT measurement is found compared to the AFM measurement of a nanostructured step-like surface. The capability of the simple system for the surface topography is further illustrated by the determination of the electrode thickness of semiconductor microelectronics. By analysing the phase change upon reflections and the optical path lengths during the measurement, the step-like structure and the sandwich configuration can be revealed from the measured FF-OCT surface maps. The usefulness of the simple system is presented in the surface topography of PMMA models. It is demonstrated that the areal refractive power can be obtained by analysing the 2-D curvature of the FF-OCT measured surface map, which is useful in the identification of surface irregularity.

Optical coherence tomography (OCT) serves as a non-destructive and non-invasive technique that is capable of imaging the inner structure of optical scattering samples with a high spatial resolution and deep penetration depth. Full-field time-domain OCT (FF-TD-OCT) is an extension of time-domain OCT (TD-OCT) which uses a two-dimensional (2D) detector to capture a series of en-face images to reconstruct the inner structure of samples in three-dimension (3D). In pharmaceutical industry, the pellet or tablet coating performs an important role in the release of active pharmaceutical ingredients (API) and controlling the desired API absorption rate in human body. Therefore, the accurate evaluation of coating thickness is vital to the pharmaceutical coating process. Our FF-TD-OCT system was developed in this research to image the pharmaceutical coating of small size pellets with a high axial resolution of 3.9μm and lateral resolution of 4.4μm. We characterized two pellet samples: a two-layer pellet with one clear coating layer and one drug-loaded layer, and a three-layer pellet with one clear coating layer and two drug-load layers. The mean thickness of a two-layer pellet was precisely determined automatically as 39.7±7.3μm and 49.1±7.0μm for the outer and inner layers respectively. The mean thickness of a three-layer pellet were 26.5±2.3μm, 20.6±3.4μm and 57.3±7.2μm respectively. In addition, the particles in the drug-loaded layer can be clearly resolved form the cross-section image. The precise and power information of the human corneal surface is of significant benefit in corneal corrective surgeries. Our developed FF-TD-OCT was combined with an average back-vertex focal length and average powercalculation algorithm in order to measure and calculate the individual power of the corneal surface. Meanwhile, the angle of incident light was considered as an important parameter and the errors introduced by the paraxial approximation was reduced. We managed to measure six formalin-fixed and two fresh corneas and map the surface power information of them. In addition, the cross-section image of cornea generated from our FF-TD-OCT system showed its structure including epithelium, Bowman’s layer and stroma clearly and the features of the stroma. For automotive paint system, the metallic flakes in base coat has a significant effect on the appearance of automotive bodies. Precise evaluation of the properties of these flakes is important in the automotive painting system in the purpose for quality assurance. Our FF-TD-OCT system was combined with a 3D variational segmentation method to measure and segment the individual flakes within the base coat of automotive paint system in 3D for the first time. The properties of flakes, including number, size and orientation in 3D space, were precisely calculated, which cannot be achieved by current commercial methods.

La rétine, en tant que fenêtre sur le cerveau, est directement connectée au système nerveux central. L'imagerie rétinienne in vivo est prometteuse pour fournir des signes de maladies neurodégénératives à un stade précoce avec une imagerie cellulaire de haute résolution. Cependant, la présence de mouvements oculaires en 3D et d'aberrations oculaires dégrade le rapport signal/bruit et la résolution de l'image. La technique de tomographie par cohérence optique plein champ dans le domaine temporel (TDFFOCT) a démontré une résolution 3D élevée et une faible complexité du système pour l'imagerie rétinienne. Un avantage de la TDFFOCT est que la résolution latérale est presque deux fois supérieure à celle des systèmes d'imagerie standard. Néanmoins, la TDFFOCT présente une sensibilité relativement faible, qui peut être notablement compromise par les aberrations oculaires et les mouvements axiaux de la rétine. En particulier, ma thèse se concentre sur la conception et la mise en œuvre de systèmes avancés de TDFFOCT avec une sensibilité de détection améliorée pour l'imagerie rétinienne in vivo. Mon manuscrit de thèse est divisé en trois parties principales. Partie 1 est une partie introductive comprenant trois chapitres. Le Chapitre 1 est consacré à la présentation de l'œil humain d'un point de vue biologique et médical et introduit les principales perturbations oculaires. Le Chapitre 2 présente l’étendue des systèmes d'imagerie clinique et de recherche pour l'imagerie de la rétine in vivo, où l'optique adaptative (AO) a été largement utilisée. Enfin, le Chapitre 3 débute par la présentation de la technique OCT, en mettant l'accent sur la TDFFOCT. La TDFFOCT surpasse les autres systèmes d'imagerie rétinienne avancés grâce à sa haute résolution 3D, un grand FOV et une conception de système compacte, permettant des applications cliniques. Cependant, la faible sensibilité du système TDFFOCT le rend difficile pour l'imagerie rétinienne interne. La Partie 2 expose la caractérisation des performances du système TDFFOCT pour l'imagerie rétinienne in vivo. Dans le Chapitre 4, je présente ma mise en œuvre d'une TDFFOCT clinique à l'Hôpital National d'Ophtalmologie Quinze-Vingts à Paris. Cependant, il reste difficile d'imager une grande population ou les couches rétiniennes internes, principalement en raison des mouvements oculaires et des aberrations oculaires qui dégradent la sensibilité. Après cela, le Chapitre 5 se concentre sur la caractérisation du mouvement axial de la rétine pour faciliter une meilleure conception d'un système de suivi rétinien. En ce qui concerne l'impact des aberrations oculaires, le Chapitre 6, propose une méthode pour étudier les performances de la TDFFOCT en présence de diverses aberrations oculaires. Ayant examiné deux paramètres principaux (mouvement oculaire et aberrations oculaires) qui dégradent la sensibilité de détection dans la TDFFOCT, des solutions pour relever ces deux défis sont également proposées, concernant 1) la fréquence de boucle requise pour un suivi axial précis de la rétine, 2) le gain du rapport signal/bruit et de la résolution par correction des aberrations pour différentes populations. Enfin, je présente la mise en œuvre de ces solutions dans le système TDFFOCT pour améliorer la sensibilité. La Partie 3 présente la conception et la mise en œuvre du système TDFFOCT avancé avec une sensibilité améliorée. Dans cette partie, j'ai mis en œuvre trois nouvelles fonctionnalités principales : • Chapitre 7 présent la conception d'une approche AO sans capteur de front d'onde pour l'imagerie rétinienne in vivo en milieu clinique. • Pour explorer la plus grande sensibilité de la TDFFOCT, j'ai conçu et mis en œuvre un TDFFOCT AO avec un capteur de front d'onde dans le Chapitre 8. • Le Chapitre 9 démontre comment améliorer les performances de suivi axial concernant les mouvements de la rétine
Retina serves as a window to the brain, and in-vivo retinal imaging is promising to give signs of neurodegenerative disease at the early stage with high-resolution cellular imaging. However, the presence of 3D eye motion and ocular aberrations degrades the image signal-to-noise ratio (SNR) and resolution. The technique of time-domain full-field optical coherence tomography (TDFFOCT), developed by Claude Boccara’s team has demonstrated exhilarating performances for retinal imaging with 3D high resolution and low system complexity. One key advantage of TDFFOCT is that the lateral resolution is nearly twice of the standard imaging system, and it is more robust to the low-order symmetric aberrations, such as defocus and astigmatisms, which account for approximately 92% of wavefront error in ocular aberrations. However, TDFFOCT exhibits relatively low sensitivity, which can be notably compromised by ocular aberrations and axial retinal motions. In particular, my PhD focuses on the design and implementation of advanced TDFFOCT systems with enhanced detection sensitivity for in-vivo retinal imaging, especially the visualizations of inner retina features with low reflectivity. My thesis manuscript is divided into three main parts. Part 1 is an introductory part comprising three chapters. Chapter 1 is dedicated to presenting the human eye from a biological and medical perspective. This highlights the imaging requirements for achieving high-resolution in-vivo retinal imaging. Chapter 2 presents the zoology of clinical and research imaging systems to image retina in vivo, where adaptive optics (AO) have been widely applied. This chapter also highlights the limitations in current AO imaging systems, mainly due to a limited field of view (FOV) and complex system design, hindering its applications in clinics. Finally, Chapter 3 starts with the OCT technique, with a focus on time-domain FFOCT. TDFFOCT outperforms other advanced retinal imaging systems, with high 3D resolution and large FOV and a compact system design, enabling clinical applications. But the low sensitivity of TDFFOCT system makes it challenging for inner retinal imaging. Part 2 is based on the characterization of the performances of the TDFFOCT system for in-vivo retinal imaging. In Chapter 4, I show my implementation of a clinical TDFFOCT which I have optimized and installed at the Quinze-Vingts National Ophthalmology Hospital, following by in-vivo retinal imaging for patients. But it’s still challenging to image a large population or inner retinal layers mainly due to eye motion and ocular aberrations degrading the sensitivity. Following this, Chapter 5 focuses on the characterization of retinal axial motion to facilitate a better design of retinal tracking system. Regarding the impact of ocular aberrations, a novel method is proposed in Chapter 6 to investigate the performances of TDFFOCT under various ocular aberrations. Having investigated two main parameters (eye motion and ocular aberrations) that degrade the detection sensitivity in TDFFOCT, the solutions to address these two challenges are also proposed regarding the loop rate required for a precise axial retinal tracking and the gain of the SNR and resolution by aberration correction for different population. Next, I will implement these solutions into TDFFOCT system to enhance the sensitivity. Following this, Part 3 shows the design and implementation of the advanced TDFFOCT system with enhanced sensitivity. In this part, I have implemented three main new features: • Chapter 7 focuses on the design of an efficient SAO approach for in-vivo retinal imaging in clinics. • To explore the highest sensitivity in TDFFOCT, I have designed and implemented a sensor-based AO TDFFOCT in Chapter 8. • Chapter 9 demonstrates how to improve the axial retinal tracking performance to facilitate more efficient frame accumulations for image averaging to improve image signal-to-noise ratio

Ce manuscrit de thèse de doctorat présente la conception et la réalisation d’un système d’imageriepour le diagnostic précoce des pathologies de la peau. Un diagnostic précoce permet de réduire lesactes chirurgicaux inutiles. Il est important de mettre en avant que seulement 20% des pathologiesfaisant office d’une opération chirurgicale, sont malignes. De plus, les pronostics de l’année 2015avançaient trois millions de nouveaux cas de cancer de la peau diagnostiqués aux ´ Etats-Unis. Basésur la tomographie par cohérence optique à balayage en longueur d’onde et une configuration pleinchamp et multi-canaux, le système d’imagerie médicale est capable d’imager en volume les couchesinternes de la peau et donc de fournir un diagnostic médical pour le professionnel de santé. Pourune fabrication en série du système portatif, les composants optiques sont micro-fabriqués sur dessubstrats et assemblés verticalement. Ces micro-composants optiques requièrent une caractérisationspécifique. Pour cela, deux systèmes ont ainsi été développés pour estimer leurs performancesoptiques. Ce travail a été réalisé dans le cadre du projet Européen VIAMOS (Vertically IntegratedArray-type Mirau-based OCT System)
The manuscript concerns the optical design and the development of a non-invasive new imagingsystem for the early diagnosis of skin pathologies. Indeed, an early diagnosis can make the differencebetween malignant and benign skin lesion in order to minimize unnecessary surgical procedure.Furthermore, prognosis for the year 2015 was that more than three millions new skin cancer caseswill be diagnosed in the United States. Based on the swept source optical coherence tomographytechnique in full-field and multiple channels configuration, the imaging system is able to perform avolumetric image of the subsurface of the skin, and thus can help in taking a better medical decision.Furthermore, for a batch-fabrication of the hand-held device, micro-optical components were made atwafer-level and vertically assembled using multi-wafer bonding. This miniaturized system requiresspecific characterization. Thus, two systems were also developed for imaging quality evaluation ofmicro-optical elements. This work has been supported by the VIAMOS (Vertically Integrated ArraytypeMirau-based OCT System) European project

Cette thèse décrit le design, la simulation et la fabrication d’une matrice 4x4 de micro-miroirs actionnée verticalement et munie d’un capteur de position. Le micro-scanneur vertical a pour vocation à être intégré au sein d’un micro-interféromètre de Mirau de type matriciel, réalisé àbase de composants micro-optiques fabriqués grâce à des méthodes collectives. Le mouvement du micro-scanneur, développé dans cette thèse, génère un signal de référence utilisé pour l’implémentation de l’interférométrie à modulation de phase dans un système de tomographie par cohérence optique (OCT). Dans un premier temps, la thèse introduit le besoin d’un système d’imagerie adapté pour la détection précoce des cancers de la peau et établit les spécifications optiques requises par cette application. A partir de ces spécifications, le design du système OCT basé sur le micro-interféromètre de Mirau est présenté. En parallèle, l’état de l’art des technologies de micro-actionnement est décrit et un actionnement électrostatique à base de peignes interdigités est choisi pour actionner et lire la position de la matrice de micro-miroirs. En effet ce type d’actionnement bénéficie d’une bonne compatibilité avec le design du micro-interféromètre de Mirau. Dans un second temps, le cœur de la thèse expose le développement du micro-scanneur vertical, c.à.d le design et les simulations ainsi que la fabrication et la caractérisation
This thesis describes the design, simulation and fabrication of a vertically actuated 4x4 array ofmicromirrors with embedded position sensing function. The vertical microscanner is meant to beintegrated within an array-type Mirau microinterferometer realized with optical microcomponentsfabricated using collective techniques. The microscanner, developed in this thesis, provides areference signal that is used for the implementation of phase modulation interferometery in an opticalcoherence tomography (OCT) system. This thesis first introduces the need for adapted imagingsystems for the early diagnosis of skin cancer and establishes the optical specifications requiredby this specific application. Based on these specifications, the design of the OCT system based onthe Mirau microinterferometer is presented. In parallel, the state of the art of the microactuationtechnologies is discussed and comb drive electrostatic actuation is chosen, for its compatibilitywith the design of the Mirau microinterferometer, to actuate and sense the position of the array ofmicromirrors. Then, the core of the thesis deals with the development of the vertical microscanner,i.e. its design and simulations, its fabrication and its characterization

Background: Optical coherence tomography (OCT), multifocal electroretinography (mfERG) and full-field electroretinography (ffERG) give important information about retinal structure and function. Purpose: To collect normative data of macular Cirrus Spectral domain (SD)-OCT assessments and of mfERG measurements of healthy children (papers I and II). To assess the macular thickness with Cirrus SD-OCT and the retinal function with ffERG in 6.5-year-old children born extremely preterm and in children born at term (papers III and IV). Methods: Study participants aged 5-15 years and living in Uppsala County were randomly chosen from the Swedish Birth Register (papers I and II). In papers III and IV, the study participants consisted of children born extremely preterm and children born at term – all were aged 6.5 years. In paper III, the children were living in Stockholm and Uppsala health care regions and, in paper IV, in Uppsala health care region only. Macular thickness was assessed with Cirrus SD-OCT and macular function with mfERG, using the Espion Multifocal system and DTL-electrodes. The retinal function was assessed with ffERG and DTL-electrodes, using the Espion Ganzfield system. Results: Altogether, 58 children participated in paper I and 49 children in paper II. In paper I, the repeatability and reproducibility of the OCT assessments were good. In paper II, the results of the mfERG measurements were in accordance with retinal cone density and there were no significant differences between the right and left eyes. In paper III, 134 preterm children and 145 children born at term constituted the study population. The central macular thickness was significantly thicker in the preterm group than in the control group. Within the preterm group, gestational age (GA), former retinopathy of prematurity (ROP) and male gender were all important risk factors for an increased macular thickness. In paper IV, 52 preterm children and 45 control children constituted the study population. Significantly lower amplitudes and prolonged implicit times of the combined rod and cone responses, as well as of the isolated cone responses, were found in the preterm group when compared with the control group. In paper IV, there was no association between GA, ROP or male gender and the ffERG assessments. Conclusion: Normative data of Cirrus SD-OCT and mfERG assessments were reported. The results of the assessments were reliable. Children aged 6.5 years, born extremely preterm, had a significantly thicker central macula and both rod and cone function were significantly reduced in comparison to children born at term. ROP had an influence on retinal structure but not retinal function in the present cohorts. Our results suggest that retinal development is abnormal in children born extremely preterm. Long-term follow-up studies are necessary in order to evaluate the functional ophthalmological outcome in this vulnerable population of children growing up today.

L’OCT plein champ est une technique d’imagerie permettant de faire de la microscopie des milieux diffusants à une profondeur donné à l’aide d’un système interférométrique. L’un des principaux objectifs qui ont motivé le développement de l’OCT plein champ ces dernières années a été de pouvoir fournir au chirurgien, durant les opérations des images similaires aux coupes histologiques mais en temps réel. Actuellement, les diagnostiques effectué à partir des images d’OCT plein champs donne de bon résultats, notamment dans le cas de certaine pathologie mammaire. Cependant, la différence majeure entre les coupes histologiques et les images d’OCT plein champs est qu’en OCT plein champ le contraste est uniquement un contraste morphologique et que dans certain cas, ce contraste seul ne suffit pas pour faire le diagnostic. Au cours des travaux de recherche décrits dans cette thèse nous nous sommes attachés à développer de nouvelles approches tomographiques, conduisant à de nouvelles formes de contraste susceptible d’enrichir les images d’OCT plein champ pour une meilleure qualité du diagnostic. Plus précisément, nous avons travaillé sur la détection optique et photo-thermique de nanoparticules d’or par OCT plein champ et la cartographie des propriétés mécaniques par trois méthodes dans le but de développer de nouveaux types de contrastes qui pourront permettre d’améliorer le diagnostic
Full Field OCT (FF-OCT) is an imaging technic use to do microscopy inside scattering media at a given depth using an interferometric setup. One of the main objectives that motivated the development of FF-OCT was to provide during surgery to the surgeon images similar to histological slices but in real time. Currently, diagnostic made from FF-OCT images gives good results, especially in the case of some breast disease. However, the major difference between the histological and FF-OCT is that FF-OCT has only a morphological contrast and in some cases, this contrast is not enough to make the diagnosis. In the research described in this thesis we are committed to developing new tomographic approaches, leading to new forms of contrast may enhance images of FF- OCT for a better quality of diagnosis. Specifically, we worked on the optical and photothermal detection of gold nanoparticles by FF-OCT and the mechanical properties mapping by three methods in order to develop new types of contrasts that will help improve the diagnosis

Cette thèse traite de l’étude et du développement d’un système d’optique adaptative pour la tomographie par cohérence optique plein champ (AO-FFOCT en anglais) appliquée à l’imagerie haute résolution de la rétine. L’analyse de l’effet des aberrations géométriques sur les performances en FFOCT a montré que pour une illumination spatialement incohérente, la résolution transverse est insensible aux aberrations et ne fait que diminuer le niveau du signal. Comme ce sont des aberrations de bas ordres comme la myopie et l’astigmatisme qui prédominent pour l’œil humain, une méthode d’optique adaptative avec une configuration sans conjugaison qui utilise une correction de front d’onde en transmission est suggérée, puis appliquée à la correction de ces ordres afin de simplifier le système d’AO-FFOCT. Des corrections de front d’onde sont effectuées sans analyseur de surface d’onde, en utilisant le niveau du signal de FFOCT comme métrique. Des expériences avec des échantillons diffusants et un œil artificiel sont menées pour démontrer la faisabilité d’un système d’AO-FFOCT conçu pour la correction d’aberration. Afin de résoudre les problèmes posés par les mouvements oculaires et de compenser en temps réel la différence de chemin optique entre les deux bras de l’interféromètre, l’instrument de FFOCT est couplé à un système d’OCT spectral. Avec cette combinaison de systèmes, l’imagerie FFOCT in vivo cellulaire de la rétine à haute résolution a été réalisée pour la première fois sur l’œil humain
This thesis follows the study and development of an adaptive optics full-field optical coherence tomography (AO-FFOCT) system, aiming for high resolution en face human retinal imaging. During the quantification of the effects of geometrical aberrations on the FFOCT system performance, it is shown that, with spatially incoherent illumination, the lateral resolution of FFOCT is insensitive to aberrations, which only cause the FFOCT signal reduction. Since low order aberrations like myopia and astigmatism dominate in human eye, a non-conjugate AO configuration by using transmissive wavefront corrector is suggested and applied for low order aberrations correction to simplify the AO-FFOCT system. Wavefront corrections are done with a wavefront sensorless method by using FFOCT signal level as the metric. Experiments with scattering samples and artificial eye model are conducted to demonstrate the feasibility of the customized AO-FFOCT system for aberration correction. In order to resolve the eye motion effects and employ real-time matching of the optical path lengths of the two interferometric arms in FFOCT, a system combination of traditional spectral-domain OCT (SDOCT) with FFOCT is adopted. With this combined system, high resolution FFOCT cellular retinal imaging is achieved in human eye in vivo for the first time

Ce manuscrit de thèse présente l’étude des techniques de super-résolution latérale en nanoscopie optique, qui est une des nouvelles techniques d'imagerie haute résolution, aujourd'hui largement utilisée en biophysique et en imagerie médicale, pour imager et caractériser des nanostructures, tout en conservant les avantages de l'imagerie optique en champ lointain comme un vaste champ, la visualisation et l’analyse en temps réel…Un des défis futurs de la microscopie 3D super-résolue est d’éviter l’utilisation des marqueurs fluorescents. La microscopie interférométrique fait partie des techniques d’imagerie 3D sans marquage permettant la détection de nanostructures. Pour améliorer le pouvoir de détection de ce système optique, un premier protocole de traitement d’images a été développé et implémenté, permettant ainsi de révéler des structures initialement non mesurables. Puis, pour améliorer la résolution latérale du système, une nouvelle technique combinant l’interférométrie et le principe du nano-jet photonique a été développée permettant l’observation d’objets de taille inférieure à la limite de diffraction de l’instrument optique
This manuscript presents the study of the lateral super-resolution techniques in optical nanoscopy, which is a new high-resolution imaging method now widely used in biophysics and medical imaging, to observe and measure nanostructures, with the advantages of far field optical imaging, such as a large field of view, visualization and analysis in real time…One of the future challenges of 3D super resolution microscopy is to avoid the use of fluorescent markers. Interferometric microscopy is a 3D label-free imaging technique enabling the detection of nanostructures. To improve the detection capability of this optical system, a first version of a protocol composed of image processing methods was developed and implemented, revealing structures initially unmeasurable. Then, to improve the lateral resolution of the system, a new technique combining interferometry and the principle of the photonic nano-jet has been developed, thus allowing the observation of objects of a size smaller than the diffraction limit of the optical instrument

碩士
國立臺灣大學
光電工程學研究所
104
In this thesis, Mirau-based full-field optical coherence tomography system using homemade Ti:sapphire crystal fiber amplified spontaneous emission (ASE) as light source was demonstrated. The lateral and axial resolution of the system are 1.25 μm and 1.3 μm in tissue, respectively. High-resolution in-vivo human skin images acquired from various locations such as dorsal forearm, palm, finger and wrist were demonstrated. Besides the morphology, dermal-epidermal junction, sweat gland, pore and vessel could be found out. The stratum corneum thickness was compared between above locations. It is about 300 μm, the thickest, at thumb and about 4 μm, the thinnest, at wrist. Complementary to the morphology extracted with OCT system, orthogonal polarization spectral imaging (OPSI) was used to study the dynamic properties of red blood cells (RBC). Due to the absorption difference, RBC under OPSI has different appearance. At 560 nm, higher absorption, RBC is a black dot. At 780 nm, lower absorption, the boundary of RBC can be seen, but not that obvious compare with the circumstances of 560 nm. Finally, RBC velocity calculation was demonstrated with particle image velocimetry (PIV) method using cross-correlation algorithm. Using regional standard deviation, RBC flowing region can be found out. The RBC velocity of capillary is about 30 - 80 μm/s, and that in small vessel is at the range of 70 – 100 μm/s. Skin morphology acquired from OCT system combined with dynamic properties studied by OPSI and PIV may provide assistance for skin diseases related with structural change and RBC velocity change.

碩士
國立臺灣大學
光電工程學研究所
102
Optical coherence tomography (OCT) has become one of the most important techniques in biomedical imaging realm. Its axial resolution is determined by central wavelength and bandwidth of the light source. Due to this reason, we made the Ce3+:YAG single-cladding crystal fiber, which generate amplified spontaneous emission centered at 560 nm with bandwidth of 95 nm. And we use it as the light source to demonstrate a full-field OCT with axial resolution of 1.468 μm and lateral resolution of 1.28 μm. In this thesis, we discuss about axial-first and lateral-first image processing method. In the section of lateral-first method, we further discuss the influence on resolution when tuning the range of band-pass filter. Then we give a conclusion that the upper limit of filter should be higher than the maximum spatial frequency sampled by lateral resolution, and the lower limit of filter only need to filter out the DC component. In this condition, both processing methods perform comparable resolution. Therefore, the instantaneity of OCT images could be improved in lateral-first image processing method. We use full-field OCT to scan the single cell specimen and find out that the most suitable reflectivity of reference mirror is about 6%. In order to measure the deeper region of specimen, we design a component to compensate dispersion effect named adjustable dispersion compensator (ADC). We also build a mathematical model to describe the operation of ADC and successfully verify its validity from experimental result. Finally, we demonstrate that the quality of 330-μm-deep single cell OCT image can be improved by the help of ADC and shows the application potential of this homemade component.

碩士
國立臺灣大學
光電工程學研究所
106
Accurate diagnosis of corneal pathology and morphological identification of different corneal layers require clear delineation of corneal three-dimensional structures and en face or cross-sectional imaging of palisade of Vogt (POV), neovascularization or corneal nerves. Here we report a prototype of full-field optical coherent tomography (FF-OCT) system with isotropic sub-micron spatial resolution in the en face and cross-sectional views. It can also provide three-dimensional reconstructed images and a large field of view by stitching tomograms side by side. We validated the imaging power of this prototype in in vivo rat and rabbit eyes, and quantified anatomical characteristics such as corneal layer thickness, endothelial cell density and the intensity profile of different layers. The mean thicknesses of rat corneal epithelium, Bowman-to-stroma, Descemet’s membrane, and endothelium were 54.1±4.0 μm, 97.4±5.0 μm, 5.9±0.8 μm, and 1.7±0.3 μm, respectively. The cell density of rat polygonal endothelial cells was quantified to be 2211±166 per mm2. This FF-OCT delineated the ridge-like structure of POV, corneal nerve bundles, and conjunctival vessels in rat eyes. It also clearly identified the vessel walls and red blood cells in rabbit model of corneal neovascularization (NV). The findings provided by this FF-OCT are expected to facilitate corneal disease diagnosis and treatment.

碩士
逢甲大學
光電研究所
98
This study provides a full-field optical coherence tomography (OCT) by using a QHQ phase shifter. The image quality of the proposed OCT is less affected by the phase shift error. The OCT system is composed of a superluminescent diode, polarization Linnik interferometer, QHQ phase shifter, and CCD. Four interferometric images corresponding to phase shifts of 0°, 90°, 180° and 270° are recorded. The phase shift is introduced between the measurement and reference beams by rotating the half-wave plate in the OHQ phase shifter. A two-dimensional cross-sectional image of a sample to be tested is then retrieved from the four images by using the four-step algorithm. The capability of the presented OCT is verified by imaging a coin covered with transparent and diffusion films, an onion, and a baby shrimp.

碩士
元智大學
光電工程學系
104
Conventional handheld skin cameras are suitable for two-dimensional (2D) inspection of shallow skin. With noninvasiveness and high resolution, optical coherence tomography (OCT) has become a popular medical imaging technology. Among OCT schemes, full-field optical coherence tomography (FF-OCT) is suitable for rapid three-dimensional (3D) imaging, as it uses a 2D imaging device for pixel processing of a sample plane. Conventionally, the use of low-coherent infrared light sources in OCT resulted in monochromatic images. As light emitting diode (LED) has advantages of wide bandwidth and long lifetime, RGB LEDs were applied in an FF-OCT system in this study to establish full-color skin imaging system by overlapping the respective red (R), green (G) and blue (B) tissue images. In this thesis, the experiment results such as full-color melanin images of a guppy, 3D skin images of a hog, and 3D skin images of human face show the potential of RGB LED sources in FF-OCT applications for dermal tissue imaging. The skin imaging system shown in the study could be miniaturized to a handheld model. Due to its non-invasiveness, high spatial resolution and the potential for full-color or 3D imaging, this imaging system can further be applied in the fields of skin science, dermatology and cosmetology.

碩士
逢甲大學
光電研究所
99
This research presents a planar liquid-crystal (PLC) phase shifter for full-field optical coherence tomography (FF-OCT). The proposed device consists of three PLC half-wave plates sandwiched between two quarter-wave plates. Compared with traditional achromatic phase shifters with a rotating wave plate or polarizer, the PLC device can quickly generate phase shifts without mechanical vibrations. Three different drive methods are applied to the PLC cells, and three different phase shifts, -120°, 0°, and 120°, are introduced between the measurement and reference beams in the FF-OCT system. A tomographic image is retrieved from three phase-shifted interferograms by using three-step algorithm. The feasibility of the proposed technology is demonstrated by imaging a coin, a coin covered alternately with transparent and diffusion films, and an onion.

博士
國立臺灣大學
光電工程學研究所
106
In the past, the quantitative analysis of biological cell is to establish information via two-dimensional (2D) images. Now, raw volumetric images were acquired through a full-field optical coherence tomography (FF-OCT) system with submicron resolution—i.e. 0.8 µm in lateral and 0.9 µm in axial direction. With the tomographic volumetric image data, more comprehensive information could be acquired. A random rayburst sampling (RRBS) framework was developed to detect the nucleus and cell membrane boundaries in three-dimensional (3D) space for determining the volumetric nuclear-to-cytoplasmic (N/C) ratio and morphology of a single cell. The design principle of this RRBS framework is to obtain the stable and objective results through random sampling. Therefore, the RRBS framework could reducing the inconvenience of manual operation and avoiding human intervention. The RRBS framework was insensitive to the selection of seeds and voxel noise. The relative standard deviation of the N/C ratio between different randomly selected seed sets was only 2%. Compared with the fluorescence confocal microscopy, the error of our algorithm is 1%. Compare with 2D images, need more than 500 exactly identical cells to obtain a precise (%RSD < 5%) volumetric N/C ratio. In this research, the volumetric N/C ratio of every single cell measured by FF-OCT could be calculated by RRBS framework. To improve computational efficiency, the graphics processing unit (GPU) is used for parallel processing. By using Nvidia GTX 1080, we have 280 times speed up for 3D bilateral filter. The RRBS framework was applied to in vivo human skin tissue with quantitative studies of the spatial distribution and morphometry of keratinocytes. The volume size of the keratinocyte nuclei became smaller from the stratum granulosum to the stratum basale. The rapid measurement of FF-OCT combined with a random rayburst sampling framework for the extraction of nuclear morphological features could serve as an effective approach to acquiring essential quantitative information for in vivo human skin tissue. Also, the spatial distribution of keratinocytes nuclei from in vivo human skin tissue were analyzed by this segmentation algorithm. The average axial and lateral diameters of keratinocyte nuclei in the stratum basale were 5.1±0.34 and 7.2±0.74 µm, respectively. The RRBS framework working with the high-resolution home-made OCT could serve as an effective approach to acquiring essential quantitative information for early stage disease diagnosis.

碩士
國立臺灣大學
光電工程學研究所
107
Construct clear delineation of corneal three-dimensional, en-face and cross sectional imaging provide more complete information to accurate diagnosis of corneal pathology and morphological identification of different corneal layers. Changes in corneal neuromorphic parameters are associated with ocular diseases. Thus, more precise technologies for analyzing innervation architecture of the cornea will be required. Here we develop Ce3+:YAG and Ti:sapphire light source full-field optical coherence tomography (FF-OCT) systems with isotropic micron spatial resolution in en-face and cross sectional views. Our systems also provide wide area scanning to reconstruct large field of view by stitching tomograms side by side. We think of a new three-dimensional concept to deal with wide range quantitative results of oblique stromal nerve. Otherwise, we applied this method on in vivo rat stromal image and got quantitative results in corneal nerve width and thickness of 17.9(2.5) and 2.2(0.2), respectively. With high spatial resolution we could quantify more reliable neuromorphic parameters of sub-basal nerve plexus. Ex vivo mouse quantitative results of corneal nerve density, diameter and tortuosity were 28.85(1.2), 1.29(0.06), 0.029(0.002) and 9.73(2.44), 1.52(0.02), 0.014(0.005), respectively, from Ce3+:YAG and Ti:sapphire system. OCT as a novel approach to assess cellular activity nondestructively. We observed different types cell colonies and nerve cells from Neuro-2A cell and limbal epithelial cell co-culture system. Moreover, we identify neurite of Neuro-2A cell and three layer structures in limbal epithelium on amniotic substrate.

碩士
國立臺灣大學
光電工程學研究所
105
Early stage diagnosis has been shown to be an efficient way to prevent the deterioration of eye diseases. In recent years, optical coherence tomography (OCT), as an important clinical imaging modality, has been widely used in clinical diagnosis in ophthalmology, helping follow the diseases progression and monitor response to therapy. Full-field optical coherence tomography (FF-OCT), a branch of OCT, has a simple setup, and its scanning speed is comparable to other fast-scanning OCT configuration, like sweep source OCT. In this work, two FF-OCT systems which both adopted amplified spontaneous emission (ASE) generated from the homemade Ti:sapphire crystal fiber as light source were demonstrated. The first system of this work is a Michelson-based FF-OCT system, which is the first FF-OCT system applied to in-vivo rat retinal and choroidal measurements in literature. The central wavelength of the light source of this system is 769.7 nm, and the 3-dB bandwidth is 163.5 nm, as a result, the system has an axial resolution of 2.1 μm in air and 1.58 μm in retinal tissue. The system revealed 5 layers of retinal image and the exquisite structure of choroid. It was found that the retinal thickness is about 250 μm, the cell size of retinal pigment epithelium is between 15 to 20 μm, and choroidal thickness is between 20 to 42 μm, and these results are consistent with these in literature. The second system employs a homemade Mirau-based FF-OCT to conduct in-vivo rat cornea measurement. The central wavelength of the system light source of is 769.9 nm, and the 3-dB bandwidth is 163.8 nm, giving the system 1.67-μm axial resolution in air and 1.21 μm in corneal tissue, and the lateral resolution is 1.12 μm. With the system, a clear corneal image with 4-layer structure can be obtained. After image processing and analysis, the corneal thickness (~160 μm), the thickness of corneal epithelium (20-30 μm), the location of the cell, the thickness of corneal endothelium (~1.4 μm) and endothelial cell density (2,998-3,217 μm) can also be calculated, and the results are consistent with these in literature. This work shows the potential of applying FF-OCT to in-vivo eye measurement. These FF-OCT systems can help academic research in current stage, and with further development, they are expected to be introduced to human trials.

碩士
國立臺灣科技大學
電子工程系
106
In this thesis, Cr:forsterite crystal fiber amplified spontaneous emission(ASE) light source was used to build a Mirau-based full-field optical coherence tomography system using two 520-nm laser diode to pump the different axis Cr:forsterite crystal fiber. The C-axis fiber can generate 3-mW ASE light centered at 900 nm with a bandwidth of 250 nm was generated. The B-axis fiber generated ASE light centered at 900nm with a bandwidth of 200um. The mirau-based full-firld OCT has an axial resolution of 1.2um and a lateral resolution of 0.98um. Skin biopsy is the gold standard for doctor to diagnosis skin cancers. There are several ways to do the skin biopsy such as punch biopsy, shave biopsy, incisional biopsy, incisional biopsy. These methods are all invasive that make the patient bleed and leave scar. Optical coherence tomography is one of the techniques in the biomedical imaging system. It is a noninvasive, label-free and 3D imaging method. In this system, we can acquire in-vivo skin image and get high-resolution 3D volume structure. In this thesis, we especially improve the design of our Mirau device to make interference efficiency increase from 7.64% to 22.35% and effective increase the noise signal ratio and obviously improve the quality of image.

碩士
國立臺灣大學
光電工程學研究所
106
Complete characterization of biological tissue requires probing the morphological and biochemical information. Optical coherence tomography (OCT) excels at presenting three-dimensional (3-D) images of tissue microstructure but lacks biomolecular information, while Raman spectroscopy is capable of providing tissue biochemical composition with superb specificity, but cannot gather 3-D microstructure. In order to obtain the complete biological sample information, in this work, we develop an integrated full-field OCT (FF-OCT) and Raman spectroscopy system capable of acquiring 3-D morphology and biochemical composition from biological samples, and demonstrate the capabilities of this system with in vitro cell line, as well as ex vivo human skin tissue. In the in vitro melanoma cell line measurement, first, our FF-OCT system is used to construct the 3-D image of melanoma cells and acquire their position in space. Then, we utilize this 3-D image to guide the acquisition of Raman spectrum from a localized melanoma cells. The experimental results show that 3-D OCT images can truly present the 3-D morphology, size (10–15 μm) and specific position in space of cells. Furthermore, the characteristic Raman peaks of melanoma cells (1326 cm-1、1469 cm-1、1660 cm-1、2939 cm-1) are indeed detected by Raman spectroscopy, which is consistent with the results found in literature. In the ex vivo experiment, we measure the melanoma biopsy and normal skin biopsy, and try to discriminate between them by the integrated FF-OCT and Raman spectroscopy system. The experimental results show that it cannot simply distinguish between melanoma biopsy and normal skin biopsy by the OCT image, but there is also a difference in the Raman spectrum at 1326 cm-1, enabling us to differentiate melanoma skin tissue and normal skin tissue by this Raman signal. This work shows the potential of applying integrated FF-OCT and Raman spectroscopy system to biological tissue measurements. By using this system, not only the 3-D tissue microstructure can be obtained, but also the biochemical composition, resulting in a more complete analysis of biological samples, which provides better assistance for the diagnose of diseases.

碩士
國立臺灣大學
光電工程學研究所
103
Skin biopsy is the gold standard for doctors to diagnosis skin cancers. There are several ways to do a skin biopsy such as punch biopsy, shave biopsy, incisional biopsy. These methods are all invasive that make the patient bleed and leave scar. Optical coherence tomography (OCT) is one of the most important techniques in biomedical imaging realm. It is a noninvasive, label-free, and 3D imaging method. In this thesis, a homemade Ti:sapphire crystal fiber amplified spontaneous emission (ASE) light source was used to build a Mirau-based full-field optical coherence tomography system. Using two 520-nm laser diodes to pump the Ti:sapphire crystal fiber, ASE centered at 770 nm with a bandwidth of 164 nm was generated. The Mirau-based full field OCT has an axial resolution of 1.67 μm in air. A 10x mirau objective was employed, and achieved a lateral resolution of 1.65 μm. Using this system, we can measure in-vivo skin and get the high resolution 3D volume structure. Utilizing the low absorption light source, the penetration depth of in-vivo skin measurement is about 300 μm. We can use this high resolution OCT image to quantitize the thickness of stratum corneum and can distinguish the dermo-epidermal junction.

碩士
國立陽明大學
生醫光電研究所
103
Optical coherence tomography (OCT) is an imaging technique that permits non-invasive and non-contact cross-sectional imaging of an object. Full-field (FF) OCT differs from time-domain and frequency-domain OCT by producing tomographic images in the en-face orientation without scanning a light beam. The entire field of the image is illuminated with a low coherence source, such as a tungsten-halogen lamp or super luminescent diode (SLD). An image sensor, such as a charge-coupled device (CCD) camera, is used for parallel acquiring of the interferometric signals. Most of existing FFOCT systems need phase modulation and synchronization apparatus to extract the tomographic image hidden in background and interference fringes, and the traditional methods are based on multiple-step phase-shifting. For the cost reduction purpose and time efficiency consideration, we have demonstrated a FFOCT system based on Linnik interferometer which is the combination of a white light emitting diode(LED) and a complementary metal oxide semiconductor (CMOS) camera. An efficient algorithms for computing FFOCT en-face images are also compared.

碩士
國立臺灣大學
醫學工程學研究所
106
The cornea is a transparent film in the front of the eye. Diseases of the cornea may lead to corneal opacification, visual impairment and even blindness. Corneal transplant surgery is a common corneal surgery, but also the high rate of success in all organ transplantations. However, there is a severe shortage of suitable cornea donors in many countries with an increased use of the old-aged population and corrective laser surgery. Therefore, the aim of this study was to construct a corneal replacement using acellular porcine corneal matrix (APCM) as a scaffold and coculturing rabbit epithelial and endothelial cells. In addition, studies of corneal reconstruction are currently analyzed and evaluated using tissue slices, but the process of tissue sectioning is quite complex and the processing time is lengthy. If non-invasive FF-OCT technology can quickly obtain biological tissue images and provide high-resolution images, it will be a great contribution in the field of research of artificial cornea. The aim of this study was to analyze the coculture of rabbit corneal epithelium and endothelium on acellular porcine cornea matrix(APCM) with full-field optical coherence tomography(FF-OCT). Results show that rabbit corneal epithelial and endothelial cells can be cocultured on APCM scaffold in vitro. According to H&E staining, epithelium layer was squeezed and detached, and so did endothelium layer. However, FF-OCT images clearly show that the corneal epithelium was composed of 5-6 layers of cells, and the corneal endothelium was a monolayer. Collectively, the study demonstrate that FF-OCT technology can quickly and efficiently achieve high-resolution analysis results of reconstructed corneal tissue.

碩士
國立臺灣大學
光電工程學研究所
107
For precise diagnosis skin and cornea pathology, en face and cross-section images are required to determine the exact pathological position, structure and cell morphology at the same time. Full-field optical coherence tomography (FF-OCT) is a kind of time-domain OCT, that scans different depth of tissues by moving PZT linear stage and collects 2-D images by camera to construct 3-D tomogram in real time. The scanning speed is depended on the frame rate of camera. Nowadays, the scanning speed of most FF-OCT are too slow, so scanning tissue will spend too much time and image quality may be worse by vibration from tissue or environment. Using high-resolution Mirau-based FF-OCT with high-speed CMOS, which frame rate is 1051 fps, 960x960x897 pixels of in vivo human skin and 960x960x574 pixels of in vivo rat cornea were scanning in 3.6 s and 2.4 s, respectively. With high spatial resolution and high-speed scanning, the structure of skin and cornea could be distinguished clearly, such like skin dermis-epidermis junction (DEJ), different corneal layers, cell morphology and boundary. The endothelium cell density of rat cornea was quantified as 2384±278 cell/mm2. Movement of single red blood cell (RBC) in microvessel could be detected, and the velocity could also be calculated as 95.54±27.82 μm/s by high-speed measurement. The high-speed Mirau-based FF-OCT not only can provide high quality lateral and axial images, but also can significantly reduce scanning time to reduce the vibration from tissue or environment.

碩士
國立臺灣大學
光電工程學研究所
104
Mirau-based full-field optical coherent tomography (Mirau FF-OCT) is a three-dimensional (3D) optical imaging system that has characteristics of non-invasive, no specific demand in preparing the sample and high spatial resolution. It has been tried to be used in clinically dermatology research lately, but part of cases are hard to distinguish lesion correctly. Fluorescence microscopy (FM) uses fluorescence dye to label biological molecule of interest, so it can only present image of specific target. We use immunofluorescence techniques to aid OCT in melanoma diagnosis. We applied labeled antibodies Cyanine Cy™3 to bind to target antibodies MART-1 for melanoma detection. We also obtained large area OCT image with exactly the same specimen, and with the help of dermatologists and pathologists circled the area on the image which showed the presence of melanoma. The H&E stain of the sample is needed as standard contrast. Our research reveals that combining fluorescence microscopy can reduce the case of misidentification dramatically. We did the experiment with two comparison methods. The first method was compared the difference in circled areas by pathologists or dermatologists between large area OCT images and H&E stains image. The second method was compared the difference in detection areas which had been involved by melanoma between combination of OCT with FM and H&E stain. We evaluated the degree of differences by using Cohen''s κ, chi-square test, phi coefficient, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). Results showed that dermatologists'' recognition as tumor area varied greatly. κ statistic, a measure of agreement, ranged from 0.74 to disagreement and phi coefficient ranged from 0.77 to 0.22. The results of diagnostic test also showed that some dermatologists recognized tumor areas very well but others did not. The circled areas by pathologists were very close. However, the results from experienced pathologists were not in agreement with H&E stain perfectly and phi coefficient was also less than 0.3. After combining FM, Cohen''s κ, phi coefficient and the results of diagnostic test are greater than 0.9. In this study, by utilizing the combination of Mirau FF-OCT with FM and immunofluorescence techniques, we obtain better results than individual methods in detecting melanoma. The agreement and correlation with H&E stain are all greater than 0.9. These demonstrate that our optical system is very close to H&E stain and has great potential on clinical application in the future.

碩士
國立陽明大學
生醫光電研究所
102
Full-field optical coherence tomography (FFOCT) is based on low coherence interferometry to extract the three-dimension backscattering images of the sample. OCT with single-point scanning required many B-scan images from scanning to reconstruct an en-face image. Thus, in order to save the time for obtaining en-face images, we used FFOCT setup in this study. In this research, a novel application is proposed that is FFOCT combines the angiographic algorithm to assess the moving backscatter in cells. In the experiment, we used the onion cells as a test sample. And then, in order to make the cells to change their morphology, we prepared the cells in hypertonic solution. We observed and captured images continuously from CCD camera. While the intensity images were captured, en-face images can be obtained after doing restructure algorithm. And then we can choose four different time-point images and calculate with angiography algorithm. At last, we analyzed and assessed the intensity of images to confirm that the cells were actually moved. For two experiment results, the images with IBDV algorithm show that cells were in contraction when they are in hypertonic solution. According the results, we confirmed the feasibility of the FFOCT system combining with the IBDV algorithm method. In the future, we hope that the novel application can be used to observe more tiny cells.

碩士
國立臺灣科技大學
光電工程研究所
104
skin biopsy is the gold standard for doctor to diagnosis skin cancers. There are several ways to do the skin biopsy such as punch biopsy, shave biopsy, incisional biopsy, incisional biopsy. These methods are all invasive that make the patient bleed and leave scar. Optical coherence tomography is one of the techniques in the biomedical imaging system. It is a noninvasive, label-free and 3D imaging method. In this thesis, Cr:forsterite crystal fiber amplified spontaneous emission(ASE) light source was used to build a Miura-based full-field Optical coherence tomography system. using two 520-nm laser diode to pumped the different axis Cr:forsterite crystal fiber, the C-axis fiber ASE centered 900 nm with a bandwidth of 250 nm was generated. The mirau-based full-field OCT has an axial resolution of 1.18 μm in tissue . The B-axis fiber ASE centered 900 nm with a bandwidth of 200 nm was generated. The mirau-based full-field OCT has an axial resolution of 1.26 μm in tissue. two kinds of water immersion objective lens were employed, when employed 40x mirau objective was achieved a lateral resolution of 0.65 um and 20x mirau objective was achieved a lateral resolution of 1.31 μm Using this system, we can measure in-vivo skin and get the high-resolution 3D volume structure. Utilizing the low absorption light source, the penetration depth of in-vivo skin measurement is about 247 μm. we can us this high-resolution OCT image to quantize the thickness of stratum corneum and can distinguish the dermo-epidermal junction. Improved of optical power of incident sample light source made the interference efficiency growth from 6.52% to 25.1% and effectively noise suppression.

碩士
國立臺灣大學
光電工程學研究所
107
Mirau-based full-field optical coherence tomography (FF-OCT) has many advantages such as high resolution, low vibration noise, and good dispersion compensation. Thus it can observe the morphology and intensity distribution of the sample but not its chemical component. Near-infrared (NIR) Raman spectroscopy can recognize the chemical institution of the sample with the advantages of low fluorescent effect and deep penetration, but not its microstructure. Therefore, measuring the five skin in vitro cell lines, including Keratinocyte cell lines, Basal cell carcinoma cell lines, Squamous cell carcinoma cell lines, Melanocyte cell and Melanoma cell lines, through combining both two techniques can acquire the microstructure, intensity distribution and chemical molecular information inside cells. Then in order to increase the accuracy and speed of the judgment in the clinical application, we use the ensemble learning algorithm of machine learning to calculate and classify them. In the in vitro experiments, the FF-OCT three-dimensional (3D) images of the five skin cell lines show the obvious protrusion on the surface of cancerous cells, that may relate to the aggressiveness of cell. And due to the smoother surface and more homogeneous internal space of the normal keratinocyte cell lines, they have lower average intensity. We utilized software to semi-automatically capture totally 283 different cells’ 3D information, including volume, compactness, surface roughness, internal average intensity, and internal intensity standard deviation. The results imply the last three features can be great parameters to distinguish between cancerous and normal cells, and the internal average intensity can be used to classify the normal melanocyte cell and keratinocyte cell. The features included compactness and volume can also distinguish different cancerous cells, but the standard deviation of these two features is too large, thus they can not be excellent indicators for classification. As a result, it needs combining the Raman spectroscopy. On the Raman spectrum of different in vitro skin cell lines, they indicate the standard deviation of the Raman spectrum from normal cells is larger than ones from cancerous cells, that may due to the biological characteristics of normal cells, that is to say, normal cells have larger variation because the cancerous cells reproduce them within one species. On the other hand, the standard deviation of Raman spectrum form cancerous cells is smaller, and also these Raman spectrums are seemingly corresponding with the literature. We can successfully classify the melanoma cells with keratinocyte-based cancerous cells / the basal cell carcinoma cells with the squamous cell carcinoma cells by six peaks and four bands (Peaks: 746, 780, 857, 1024, 1063, 1209 cm-1; Bands: 925-946, 990-1010, 1088-1130, 1281-1302 cm-1) / seven peaks and three bands (Peaks: 854, 898, 1064, 1158, 1191, 1233, 1452 cm-1; Bands: 923-946, 1007-1028, 1291-1336 cm-1) in the Raman spectrum of 600-2000 cm-1. Consequently, the classification of Raman spectrums on cancerous cells has better performance. Eventually, we employ the ensemble learning form machine learning to classify them, which also verify the above results. Decision tree method in ensemble learning has better-classified results generally. The accuracy can reach 85.9% on distinguishing normal and cancerous cell by FF-OCT features. Also, these features can distinguish every species of normal skin cell. Then Raman spectrums can completely classify three kinds of cancerous skin cell. Therefore, these techniques can distinguish these five normal and cancerous skin cell-lines very accurately and fast. And it shows the method of integrating FF-OCT, NIR Raman spectroscopy and ensemble learning can be an important diagnostic tool and direction for clinical research and application.