• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 2
  • Tagged with
  • 2
  • 2
  • 2
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Design and development of a universal handheld probe for optoacoustic-ultrasonic 3D imaging / Conception et développement d’une sonde portable universelle pour l’imagerie 3D optoacoustique-ultrasonique

Azizian Kalkhoran, Mohammad 05 April 2017 (has links)
La présente dissertation est principalement consacrée à la conception et à la caractérisation d’une sonde universelle pour l’imagerie volumétrique ultrasons-optoacoustique et le développement d’un algorithme de reconstruction adapté aux exigences physiques pour la conception du système. Les traits distinctifs de cette dissertation sont l’introduction d’une nouvelle géométrie pour les sondes manuelles ultrasons-optoacoustique et des évaluations systématiques basées sur des méthodes de pré-reconstruction et post-reconstruction. Pour éviter l’interprétation biaisée, une évaluation capable d’évaluer le potentiel de la sonde doit être faite. Les caractéristiques mentionnées établissent un cadre pour l’évaluation des performances du système d’imagerie d’une manière précise. En outre, elle permet d’optimiser les performances suivant l’objectif fixé. Ainsi, deux algorithmes de reconstruction anticipée ont été élaborés pour la conception du système OPUS (optoacoustique ultrasons) capables de produire des images avec un contraste et une résolution homogènes sur tout le volume d’intérêt. L’intérêt d’avoir de tels algorithmes est principalement dû au fait que l’analyse des données médicales est souvent faite dans des conditions difficiles, car on est face au bruit, au faible contraste, aux projections limités et à des transformations indésirables opérées par les systèmes d’acquisition. Cette thèse montre, aussi, comment les artefacts de reconstruction peuvent être réduits en compensant les propriétés d’ouverture et en atténuant les artefacts dus à l’échantillonnage angulaire parcimonieux. Afin de transférer cette méthodologie à la clinique et de valider les résultats théoriques, une plate-forme d’imagerie expérimentale a été développée. En utilisant le système de mesure développé, l’évolution d’une nouvelle géométrie annulaire parcimonieuse et sa dynamique ont été étudiées et une preuve de concept a été démontrée à travers des mesures expérimentales dans le but d’évaluer les progrès réalisés. / When the interest is in multiscale and multipurpose imaging, there exists such a will in integrating multi-modalilties into a synergistic paradigm in order to leverage the diagnostic values of the interrogating agents. Employing multiple wavelengths radiation, optoacoustic imaging benefits from the optical contrast to specifically resolve molecular structure of tissue in a non-invasive manner. Hybridizing optoacoustic and ultrasound imaging comes with the promises of delivering the complementary morphological, functional and metabolic information of the tissue. This dissertation is mainly devoted to the design and characterization of a hybridized universal handheld probe for optoacoustic ultrasound volumetric imaging and developing adaptive reconstruction algorithms toward the physical requirements of the designed system. The distinguishing features of this dissertation are the introduction of a new geometry for optoacoustic ultrasonic handheld probe and systematic assessments based on pre and post reconstruction methods. To avoid the biased interpretation, a de facto performance assessment being capable of evaluating the potentials of the designed probe in an unbiased manner must be practiced. The aforementioned features establish a framework for characterization of the imaging system performance in an accurate manner. Moreover, it allows further task performance optimization as well. Correspondingly, two advanced reconstruction algorithms have been elaborated towards the requirement of the designed optoacoustic-ultrasound (OPUS) imaging system in order to maximize its ability to produce images with homogeneous contrast and resolution over the entire volume of interest. This interest is mainly due to the fact that the medical data analysis pipeline is often carried out in challenging conditions, since one has to deal with noise, low contrast, limited projections and undesirable transformations operated by the acquisition system. The presented thesis shows how reconstruction artifacts can be reduced by compensating for the detecting aperture properties and alleviate artifacts due to sparse angular sampling. In pursuit of transferring this methodology to clinic and validating the theoretical results, a synthetic imaging platform was developed. Using the measurement system, the evolution of a novel sparse annular geometry and its dynamics has been investigated and a proof of concept was demonstrated via experimental measurement with the intention of benchmarking progress.
2

Uncertainty Estimation in Volumetric Image Segmentation

Park, Donggyun January 2023 (has links)
The performance of deep neural networks and estimations of their robustness has been rapidly developed. In contrast, despite the broad usage of deep convolutional neural networks (CNNs)[1] for medical image segmentation, research on their uncertainty estimations is being far less conducted. Deep learning tools in their nature do not capture the model uncertainty and in this sense, the output of deep neural networks needs to be critically analysed with quantitative measurements, especially for applications in the medical domain. In this work, epistemic uncertainty, which is one of the main types of uncertainties (epistemic and aleatoric) is analyzed and measured for volumetric medical image segmentation tasks (and possibly more diverse methods for 2D images) at pixel level and structure level. The deep neural network employed as a baseline is 3D U-Net architecture[2], which shares the essential structural concept with U-Net architecture[3], and various techniques are applied to quantify the uncertainty and obtain statistically meaningful results, including test-time data augmentation and deep ensembles. The distribution of the pixel-wise predictions is estimated by Monte Carlo simulations and the entropy is computed to quantify and visualize how uncertain (or certain) the predictions of each pixel are. During the estimation, given the increased network training time in volumetric image segmentation, training an ensemble of networks is extremely time-consuming and thus the focus is on data augmentation and test-time dropouts. The desired outcome is to reduce the computational costs of measuring the uncertainty of the model predictions while maintaining the same level of estimation performance and to increase the reliability of the uncertainty estimation map compared to the conventional methods. The proposed techniques are evaluated on publicly available volumetric image datasets, Combined Healthy Abdominal Organ Segmentation (CHAOS, a set of 3D in-vivo images) from Grand Challenge (https://chaos.grand-challenge.org/). Experiments with the liver segmentation task in 3D Computed Tomography (CT) show the relationship between the prediction accuracy and the uncertainty map obtained by the proposed techniques. / Prestandan hos djupa neurala nätverk och estimeringar av deras robusthet har utvecklats snabbt. Däremot, trots den breda användningen av djupa konvolutionella neurala nätverk (CNN) för medicinsk bildsegmentering, utförs mindre forskning om deras osäkerhetsuppskattningar. Verktyg för djupinlärning fångar inte modellosäkerheten och därför måste utdata från djupa neurala nätverk analyseras kritiskt med kvantitativa mätningar, särskilt för tillämpningar inom den medicinska domänen. I detta arbete analyseras och mäts epistemisk osäkerhet, som är en av huvudtyperna av osäkerheter (epistemisk och aleatorisk) för volymetriska medicinska bildsegmenteringsuppgifter (och möjligen fler olika metoder för 2D-bilder) på pixelnivå och strukturnivå. Det djupa neurala nätverket som används som referens är en 3D U-Net-arkitektur [2] och olika tekniker används för att kvantifiera osäkerheten och erhålla statistiskt meningsfulla resultat, inklusive testtidsdata-augmentering och djupa ensembler. Fördelningen av de pixelvisa förutsägelserna uppskattas av Monte Carlo-simuleringar och entropin beräknas för att kvantifiera och visualisera hur osäkra (eller säkra) förutsägelserna för varje pixel är. Under uppskattningen, med tanke på den ökade nätverksträningstiden i volymetrisk bildsegmentering, är träning av en ensemble av nätverk extremt tidskrävande och därför ligger fokus på dataaugmentering och test-time dropouts. Det önskade resultatet är att minska beräkningskostnaderna för att mäta osäkerheten i modellförutsägelserna samtidigt som man bibehåller samma nivå av estimeringsprestanda och ökar tillförlitligheten för kartan för osäkerhetsuppskattning jämfört med de konventionella metoderna. De föreslagna teknikerna kommer att utvärderas på allmänt tillgängliga volymetriska bilduppsättningar, Combined Healthy Abdominal Organ Segmentation (CHAOS, en uppsättning 3D in-vivo-bilder) från Grand Challenge (https://chaos.grand-challenge.org/). Experiment med segmenteringsuppgiften för lever i 3D Computed Tomography (CT) vissambandet mellan prediktionsnoggrannheten och osäkerhetskartan som erhålls med de föreslagna teknikerna.

Page generated in 0.0673 seconds