1,721,089 research outputs found
Modular Open-Source toolbox for optics education
Full text linkUC2 is an open-source modular toolbox for making hardware projects based on functional optical blocks. It creates a bridge between education and science by offering an alternative to the standard tools, a generic toolbox that can be used for both fields. The same basic hardware is meant to be used for teaching photonics and its applications in an interactive way and the students could later use the same system in their own work for prototyping. As a community-driven collaborative project, UC2 benefits from the experience of its early adopters and from its attractiveness for project-based courses. It is a collective effort of many researchers and students. This thesis shows the contributions of the author to the project. In this work, we present the development and testing of the hardware for the educational applications. The key aspects of open-source hardware design are evaluated and summarized into a decision matrix defining the boundary conditions of the development. Special importance is given to improving accessibility of the toolbox by providing comprehensive documentation and lowering the entry barrier. Four different educational kits, in summary called “TheBOX”, were developed, with the MiniBOX being at the stage of a production-ready prototype, optimized for mass production. The BOXes are comprehensive toolkits aimed at a specific education level or demonstration of certain experiments. These present a low-cost alternative to the commercially available systems. Additionally to the advantage in price, they offer a lower entry level and create an inviting environment for science education. They were developed in close contact with the users and the improvements were based on iteratively acquired feedback. The system was tested in many workshops and courses and provided tools for student projects
Entwicklung einer endoskopischen Fasersonde für optische Biopsien mithilfe nichtlinearer Spektroskopie
Full text linkEine frühe Erkennung und exakte Diagnostik von Krankheiten wie beispielsweise Tumoren ist unerlässlich für eine optimale Behandlung und die Verbesserung von Prognosen betroffener Menschen. Den Goldstandard für Gewebeuntersuchungen stellt bis heute die Histopathologie dar, wobei schmerzhafte Patientenbiopsien und komplexe Nachbehandlungen der Gewebeproben notwendig sind. Von der Gewebeentnahme bis zur Diagnosestellung vergeht wertvolle Zeit, in vielen Fällen müssen Biopsien mehrfach wiederholt werden, bis krankhaftes Gewebe vollständig identifiziert und lokalisiert ist. Aus diesem Grund wurden in den letzten Jahrzehnten neuartige optische Technologien entwickelt und angewendet, um Gewebeanalysen zu optimieren und eine patientenfreundlichere Behandlung zu ermöglichen. Nichtlineare, multimodale In-vivo-Bildgebung mit CARS (coherent anti-Stokes Raman scattering), SHG (second harmonic generation) und TPEF (two-photon excited auto-fluorescence) kann hier einen attraktiven Lösungsansatz bieten, welcher nichtinvasiv ist und die Detektion von mikroskopischen Gewebeveränderungen in Echtzeit ermöglicht. Die Implementierung nichtlinearer Bildgebungsverfahren in ein endoskopisches Design für schwer zugängliche Gewebestellen stellt jedoch immer noch eine technologische Herausforderung dar. Die vorliegende Forschungsarbeit präsentiert einen faserbasierten Ansatz für eine multimodale, endoskopische Sonde, wobei eine Multikernfaser für die Führung der Anregungslaser genutzt wird und speziell angefertigte GRIN-Linsen die Strahlformung im Sondenkopf übernehmen. Das Sondendesign kommt vollständig ohne bewegliche Teile oder den Einsatz von elektrischem Strom aus, so dass eine robuste und kompakte Bauform erreicht wird. Die Funktionalität der Sonde wird mit multimodalen Aufnahmen an Dünnschnitten von menschlichem Hautgewebe sowie an frischen Gewebeproben eines Hausschweins demonstriert. Anhand der Ergebnisse kann gezeigt werden, dass multimodale Aufnahmen mit der vorgestellten Sonde qualitativ vergleichbar sind mit konventionellen LSM-Aufnahmen, darüber hinaus ist eine schnellere Bildgebung möglich. Somit hat das präsentierte Sondendesign ein hohes Potential, zukünftig bei klinischen Routineuntersuchungen eingesetzt zu werden
Advanced single molecule localization microscopy for imaging cellular nuclei
Full text linkIn this PhD research, single molecule localization microscopy (SMLM) was used to image nuclear structures with a resolution down to several nanometers.The scope of this PhD research is to develop a 3D SMLM microscope which can overcome several principle limitations in imaging nuclei in 3D. The advanced improvements during this PhD research include a broad range of research subjects associated to SMLM techniques. Firstly, one of the most common problems of a super-resolution microscope is sample drift, because a small sample drift may result in artefacts and can hamper the resolution. A speckle-based method was developed to correct sample drift without changing the standard design of the SMLM setup. This drift correction method can achieve a resolution of several nanometers. Secondly, another principle problem is that commonly used organic fluorophores are restricted in their photon budget. It is often observed that the chemical structure of fluorophores change after high laser irradiance resulting in photobleaching. A patterned illumination technique was developed which allows the user to define arbitrary regions of interest for illumination with a flat-top intensity profile. Thirdly, for SMLM in particular, a carefully adjusted chemical environment in the sample is recommended to induce sufficiently blinking signals of the organic fluorophores in combination with an appropriate laser irradiance. However, such an imaging buffer can degrade over time and may not be suitable for long time imaging. Nanographene was presented as a new class of fluorophores which have blinking properties without an imaging buffer. Therefore, the nanographenes facilitate a wide range of SMLM applications including bio-imaging and material science. These advanced developments are not only for imaging nuclei, but also applicable to applications in other biological researches and in material science
High-resolution direct stochastic optical reconstruction microscopy of the human kinetochore chromatin
Full text linkThe kinetochore is a chromatin-protein complex on the chromosome centromere. Kinetochore attache chromosomes to the microtubules of the spindle apparatus, recognize attachment to microtubules, and regulate the progress of mitosis. The kinetochore protein complex is built on the periphery centromere chromatin. It seems that kinetochores react to forces of the spindle apparatus microtubules in their interior. Due to the limited resolving power of the optical microscope and limitations of electron microscopy, the spatial arrangement of kinetochore protein and chromatin complex is unresolved. In this thesis, the human periphery centromere chromatin assembled with CENP-A (kinetochore chromatin in further text), on which kinetochore complex is built, was imaged by high-resolution optical microscopy called direct Stochastic Optical Reconstruction Microscopy (dSTORM). Resolving power of less than 30 nm was achieved, and the architecture of more than 900 kinetochores in different sub-stages of interphase and mitosis was resolved. Kinetochore chromatin was formed in a rectangle between 250 and 400 nm long and between 150 and 270 nm wide. It was composed of parallel and orthogonal lines between 12 and 75 nm wide. The arrangement of kinetochore chromatin in mitosis was narrower and longer than in interphase. In interphase, subtle changes in the dimensions of kinetochore lines were measured. The mitotic toxin nocodazole disturbed kinetochore chromatin organization. The discovered change of arrangement of the kinetochore chromatin assembled with CENP-A during the cell cycle could be the physical mechanism of recognition of proper attachment and positioning of the chromosomes at the equatorial plate. Based on the discovered kinetochore structure and behavior in the cell cycle, a new shoo-lace model of assembly and function of kinetochore chromatin is suggested
Combining image scanning and stimulated emission depletion microscopy to achieve isotropic resolution
Full text linkBiologists are always interested in the molecular structures inside living cells, for example, mitochondria. But most of the time, these fine subcellular structures are randomly oriented in X, Y, and Z direction. For faithful image reconstruction, it is always desired to have an isotropic resolution improvement. But the resolution, in general, is not isotropic. In fact, it is three times worse along the axial direction than the lateral, making it challenging to study 3D structures inside the cells. Although existing techniques like 4pi and STED microscopy have been used to achieve isotropic resolution improvements, the setups are highly complex and high intensities used in the STED beam cause photobleaching and phototoxicity-induced artifacts, especially in the case of live-cell imaging. Therefore, it is crucial to develop novel microscopy approaches that simultaneously optimize both lateral and axial resolution potentially using low excitation powers. In this thesis, we realize a new combination of image scanning microscopy and π-step phase modulated depletion STED beam. The basic idea behind this is that image scanning microscopy will improve the resolution in the lateral direction and the resolution in the axial direction will be improved simultaneously by the depletion STED beam. The depletion STED beam used here is a π-step phase modulated that will deplete the fluorescence originating from the areas above and below the image plane but the fluorescence in the image plane is never quenched by the STED laser beam. Thus, their photobleaching by the STED beam can be avoided. We therefore expect much higher photon counting rates than with conventional 3D STED
Improving structured illumination microscopy by blind reconstruction and multifocus detection
Full text linkStructured Illumination Microscopy (SIM) is a super-resolution microscopy method which enables a two-fold resolution enhancement with respect to classical wide-field (WF) fluorescence microscopy. In SIM, the sample is illuminated by a pattern, typically a sinusoidal grid of light. This pattern is typically the result of interference of either two or three plane waves. The wide dissemination of SIM is limited by the difficulties arising during the necessary numerical reconstruction. Artefacts are observed in the SIM reconstructed images if the illumination pattern is distorted or unknown. Blind-SIM is a deconvolution-based reconstruction approach that enables to reconstruct both the sample and the illumination pattern. Blind-SIM is able to reconstruct partially or fully unknown illumination patterns and is therefore robust to distortions. However, so far blind-SIM was not able to process data from thick samples where the visibility of the fringes is reduced by the presence of out-of-focus light. In this work, we present the thick slice blind-SIM algorithm which reconstructs a 3D stack from a 2D measurement, thus removing the out-of-focus light. The acquisition speed in 3D SIM is limited by the axial scanning and sequential focusing. Using a multifocus detection enables to simultaneously acquire a focal series. Merging SIM excitation with multifocus detection (MF-SIM) permits to enhance the volumetric acquisition speed of 3D SIM as well as removing the need for axial mechanical scanning, a source of sample drift and vibrations. The recorded MF-SIM data does not obey the same theoretical description as the conventional scanning 3D SIM data. Hence, the classical approach for 3D SIM reconstruction cannot be applied. We developed a deconvolution-based algorithm using known 3D illumination pattern that can be applied to the reconstruction of MF-SIM data. The results demonstrate enhanced resolution in all three dimensions.Mikroskopie mit strukturierter Beleuchtung (SIM) zählt zu den hochauflösenden Mikroskopieverfahren und erreicht die doppelte Auflösung im Vergleich zur normalen Weitfeldmikroskopie. Bei SIM wird die Probe mit einem Muster – typischerweise einem Lichtgitter - beleuchtet. Dieses Muster entsteht durch die Interferenz zweier oder dreier ebener Wellen. Die nötige Datenverarbeitung der rohen SIM-Bilder führt zu Artefakten wenn das Beleuchtungsmuster unbekannt oder verzerrt ist. Blind-SIM ist ein Entfaltungsbasierter Ansatz, der sowohl die Rekonstruktion der Probe als auch die des Beleuchtungsmusters ermöglicht. Blind-SIM ist fähig, Daten mit völlig oder teilweise unbekanntem Beleuchtungsmuster zu bearbeiten und ist deswegen robust gegenüber Verzerrungen dieser Muster. Thick slice blind-SIM hingegen erzeugt einen 3D Stapel aus einer 2D-Aufnahme. Dadurch ist es in der Lage, Beiträge von Licht außerhalb der Fokusebene zu entfernen. Durch das Zweistrahl-SIM Verfahren lassen sich entweder optisch Probenschnitte erzeugen oder eine Maximierung der lateralen Auflösung erzielen. Eine Kombination mit thick slice blind-SIM ermöglicht beides gleichzeitig. Thick slice blind-SIM zeigt durch die Bearbeitung einer einzelnen ausgewählten Ebene eine vergleichbare Verbesserung des Auflösungsvermögens und der optische Sectioning Fähigkeit wie die 3D Bearbeitung eines 3D aufgenommenen SIM Stapels.
Die Aufnahmegeschwindigkeit für 3D SIM ist durch axiales mechanisches Abtasten in Verbindung mit Nachfokussierung begrenzt, wohingegen eine multifokus Detektion die gleichzeitige Aufnahme einer fokalen Serie ermöglicht. Die Kombination von SIM-Anregung und multifokus Detektion (MF-SIM) ermöglicht es, die Aufnahmegeschwindigkeit von 3D SIM zu verbessern. Das MF-SIM Verfahren macht neue Ansätze der Rekonstruktion nötig, wofür sich ein entfaltungsbasierter Algorithmus, welcher auf bekannten 3D-Beleuchtungsmustern beruht, eignet
Fast SLM-based linear and nonlinear structured illumination microscopy
Full text linkFluorescent microscopy becomes an essential tool for medical and biological investigations due to its major advantages of allowing for minimally invasive observation and rapid optical imaging. It is also a highly desirable tool to study the three dimensional interior of living specimens at a small scale. However, the resolution of optical systems is fundamentally limited by the diffraction of light, which consequently coins the development of super-resolution imaging methods. Structured illumination microscopy (SIM) is one of the super-resolution techniques. SIM provides a two-fold lateral resolution improvement for those types of samples where the fluorescence emission intensity depends linearly on the intensity of the illumination pattern. The concept of SIM is based on the Moiré effect. A structured illumination pattern is projected into the sample and high spatial frequency components of the biological sample, which are normally above the cut-off frequency of the optical transfer function and therefore lost, are then down-modulated to low spatial frequencies that reside inside the passband of the optical transfer function of the microscope. Typically, a lateral resolution of 100 nm becomes achievable in SIM. SIM is a wide-field technique and thus allows fast acquisition of large fields of view. This work discusses methods to improve the acquisition speed of SIM and to further enhance the resolution beyond the usual factor of two using nonlinear SIM (NL-SIM). Improvement of the acquisition speed is achieved by exploiting the advantages of a ferroelectric spatial light modulator (SLM) which offers fast switching of the illumination pattern, a modern sCMOS camera which provides fast readout and a novel synchronization approach between the different opto-electronical components
Multimodale chemisch-sensitive Bildgebung: Implementierung und Evaluierung neuer mikrospektroskopischer Ansätze
Full text linkInhalt dieser Arbeit ist die Entwicklung und der Test einer Analyseplattform für schwingungsspektroskopische Untersuchungen. Mit dem erhaltenen Werkzeug sollen die Grundlagen für eine Überführung und Etablierung von nicht-linearen optischen Kontrastmethoden in die klinische Anwendung gelegt werden.
Zu diesem Zweck wurde ein experimenteller Aufbau mit einem Pikosekunden-Laser und einem OPO konstruiert und verwirklicht. Mit Augenmerk auf Flexibilität, Dokumentation und Reproduzierbarkeit sowie Stabilität ist dabei eine offene Plattform entstanden, die Stimulation und Aufzeichnung mehrerer Modalitäten erlaubt. Sie wurde für die Aufzeichnung und den Vergleich vorwärts und rückwärts gestreuter Signalanteile um einen Faserschalter erweitert. Eine Leistungsregelung sorgt für einen deutlich stabileren Betrieb. Als Detektoren stehen ausgewählte Varianten von APD, PMT und Spektrometer zur Verfügung.
Zum Betrieb der integrierten Hardware und zur Erfüllung höherer Funktionen wurde eine neue Software entwickelt, die im Hinblick auf die Aufgabenstellung die Nachteile von bekannten freien oder kommerziellen Produkten beseitigt. Sie ist offen und modular konstruiert. Alle Rohdaten sind selbstbeschreibend und werden vollständig gespeichert.
Die ersten Funktionstests mit dem Mikroskop sind ebenfalls Bestandteil dieser Arbeit. Sie dokumentieren die Verwendungsmöglichkeit für hyperspektrales CARS insbesondere im Fingerprint-Bereich und darauf basierende Bildgebung. Hyperspektrales CARS wurde an einer Einzelbande von Toluol (1004 cm-1) zusammen mit einem Bandenfit zur Bestimmung der Signalanteile demonstriert, ferner an der Doppelbande von Xeloda bei 1634 cm-1 sowie in Form einer Konzentrationsreihe an Beta-Carotin (1514 cm-1). Mit Hilfe der hyperspektralen Bildgebung wird an PMMA die Funktion eines Faserschalters zur quasi-parallelen Aufzeichnung sowie an menschlicher Haut die Stabilisierung der Stimulationsleistung gezeigt
Towards clinical translation of raman spectroscopy for tumor cell identification
Full text linkIn the modern world, cancer is one of the leading causes of death, and its early diagnostics remains one of the big challenges. Since cancer starts as a malfunction on the cellular level, the diagnostic techniques have to deal with single cells. Detection of circulating tumor cells (CTCs), which are present in the patient's blood, holds promise for the future theranostic applications, as CTCs represent the actual state of the primary tumor. Raman spectroscopy is a label-free technique capable of non-destructive and chemically-specific characterization of individual cells. In contrast to marker-based methods, the CTCs detected by Raman can be reused for more specific single-cells biochemical analysis methods. This thesis focuses on technological developments for Raman-based CTC identification, and encompasses the whole chain of involved methods and processes, including instrumentation and microfluidic cell handling, automation of spectra acquisition and storage, and chemometric data analysis. It starts with a design of custom application-specific instruments that we used to evaluate and optimize different experimental parameters. A major part is software development for automated acquisition and organized storage of spectral data in a database. With the automated measurement systems and the database in place, we were able to collect about 40.000 Raman spectra of more than 15 incubated cancer cell lines, healthy donor leukocytes, as well as samples originating from clinical patients. Additionally, the thesis gives an overview of data analysis methods and provides an insight into the underlying trends of the dataset. Although the cell identification models could not reliably differentiate between individual cancer cell lines, they were able to recognize tumor cells among healthy leukocytes with prediction accuracy of more than 95%. This work demonstrated an increase in the throughput of Raman-based CTC detection, and provides a basis for its clinical translation
- …
