Background: Dynamic contrast-enhanced MRI (DCE-MRI) has the potential to produce images of physiological quantities such as blood flow, blood vessel volume fraction, and blood vessel permeability. Such information is highly valuable, e.g., in oncology. The focus of this work was to improve the quantitative aspects of DCE-MRI in terms of better understanding of error sources and their effect on estimated physiological quantities. Methods: Firstly, a novel parameter estimation algorithm was developed to overcome a problem with sensitivity to the initial guess in parameter estimation with a specific pharmacokinetic model. Secondly, the accuracy of the arterial input function (AIF), i.e., the estimated arterial blood contrast agent concentration, was evaluated in a phantom environment for a standard magnitude-based AIF method commonly used in vivo. The accuracy was also evaluated in vivo for a phase-based method that has previously shown very promising results in phantoms and in animal studies. Finally, a method was developed for estimation of uncertainties in the estimated physiological quantities. Results: The new parameter estimation algorithm enabled significantly faster parameter estimation, thus making it more feasible to obtain blood flow and permeability maps from a DCE-MRI study. The evaluation of the AIF measurements revealed that inflow effects and non-ideal radiofrequency spoiling seriously degrade magnitude-based AIFs and that proper slice placement and improved signal models can reduce this effect. It was also shown that phase-based AIFs can be a feasible alternative provided that the observed difficulties in quantifying low concentrations can be resolved. The uncertainty estimation method was able to accurately quantify how a variety of different errors propagate to uncertainty in the estimated physiological quantities. Conclusion: This work contributes to a better understanding of parameter estimation and AIF quantification in DCE-MRI. The proposed uncertainty estimation method can be used to efficiently calculate uncertainties in the parametric maps obtained in DCE-MRI.

Thesis (Ph. D.)--Ohio State University, 2008. / Title from first page of PDF file. Includes bibliographical references (p. 113-116).

Thesis (M.S)--University of Alabama at Birmingham, 2007. / Description based on contents viewed June 11, 2008; title from title screen. Includes bibliographical references (p. 31-35).

The utility of microbubble agents in a variety of diagnostic and therapeutic ultrasound techniques has been widely demonstrated, most notably in Contrast Enhanced Ultrasound (CEUS) imaging. Unfortunately, the underlying mechanisms of their response to ultrasound excitation are poorly understood, restricting the development of promising techniques, such as quantitative perfusion imaging. A significant reason for this is that current microbubble characterisation techniques suffer from one or more of the following limitations: i) large experimental uncertainties, ii) physical restrictions on microbubble response and iii) failure to provide large data sets suitable for statistical analysis. This thesis presents a new technique to overcome these limitations. A co-axial microfluidic device is used to hydrodynamically confine microbubbles through the focal region of a laser and ultrasound field. The magnitude of light scattered by isolated microbubbles during ultrasound excitation is converted to radius using Mie Scattering theory. This technique is capable of obtaining large samples (>10³/min) of microbubbles to be efficiently characterised. The response of a commercial contrast agent, SonoVue®, is first investigated for a range of ultrasound exposure parameters; frequency (2 MHz - 4.5 MHz), peak negative pressure (6 kPa - 400 kPa) and pulse length (3 cycles - 8 cycles). Second the device is used to investigate the effect of composition and fabrication on microbubble response to similar ultrasound conditions. The results demonstrate a very large variability in microbubble response independent of initial size, indicating a significant lack of uniformity of coating properties. This is further supported by quantitative fluorescence imaging and quasi-static pressure chamber measurements. The implications of the findings for CEUS imaging and the development of microbubble contrast agents are discussed, as well as the limitations and suggested improvements of the characterisation technique.

Abstract The purpose of this study was to evaluate the value of 0.23T low-field magnetic resonance imaging (MRI) and nanocolloid (NC) scintigraphy in assessing joint pathology associated with rheumatoid arthritis (RA). Fat suppression methods combined with contrast media enhancement aid in distinguishing enhancing inflamed tissue from the surrounding fat, especially in the imaging of arthritic joints. The feasibility and image quality of a phase-shift water-fat MRI method for fat suppression at low-field 0.23T open configuration MR scanner was evaluated. The technique was combined with contrast-enhanced imaging to assess the conspicuity of synovial hypertrophy in the joints of 30 RA patients. Improved conspicuity and delineation of synovitis was detected with this method. However, because of a great amount of manual post processing, future development is needed to make this method more feasible. Contrast-enhanced MRI and NC scintigraphy may provide objective and quantitative information about the inflammatory activity in arthritic joints. The value of quantitative and semiquantitative measures of inflammation derived from NC scintigraphy and low-field MRI of the wrist joint of 28 early RA patients was evaluated. Furthermore, it was investigated whether these parameters have predictive value of further erosive development during two years of follow-up. Strong correlations were detected between the NC scintigraphy and MRI measures, and these parameters were associated with laboratory markers of inflammation. During the two-year follow-up, the initial MRI and NC scintigraphy measures were closely related with the progression of wrist joint erosions. Small erosive-like bone defects can occasionally be found in wrist MRI of patients without clinically overt arthritis. The prevalence of these lesions was studied in bilateral wrist MRI examinations of 31 healthy persons. Small lesions resembling erosions were detected in 14 out of 31 subjects. Altogether 24 of the 930 wrist bones evaluated showed such lesions (3%). Thus small changes resembling erosions can be found in the wrist MRI of healthy subjects; the significance of these findings must always be interpreted with reference to the clinical picture. In conclusion, early RA patients with high local inflammatory activity, as detected by NC scintigraphy and MRI are at risk of developing further bone damage. Furthermore, in the follow-up of early RA patients, if clinically sustained response is not achieved, these methods help to identify patients who need more intensive drug treatment.

Les modifications induites par les cellules tumorales sur leur environnement ont pour but de permettre leur développement en remodelant le tissu la soutenant et en créant un nouveau réseau vasculaire (angiogenèse). Plusieurs thérapies anti-angiogéniques inhibant le développement du réseau vasculaire tumoral ont obtenu l’autorisation de mise sur le marché et sont actuellement utilisées en clinique. Ces thérapies induisent de fortes modifications fonctionnelles au sein de la tumeur mais le simple suivi de l’évolution du volume tumoral n’est pas suffisant pour rendre compte de ces modifications. L’objectif principal de la thèse a consisté à utiliser différentes modalités d’imagerie ultrasonore afin d’évaluer leur sensibilité aux modifications générées dans des tumeurs murine (carcinome colorectal et pulmonaire) au cours de plusieurs types de thérapie (chimique : cytotoxique, anti-angiogénique / physique : plasma froid, sono-sensibilisation). Les modifications de la distribution spatiale des micro-vaisseaux et leur fonctionnalité ont été caractérisées à l’aide de l’imagerie de contraste ultrasonore (CEUS), l’altération de la microstructure de la tumeur a été évaluée grâce à l’analyse spectrale des signaux radiofréquences, connu comme « quantitative ultrasound » (QUS) et enfin les variations des propriétés mécaniques des tissus tumoraux ont été mesurées en élastographie à l’aide de la technique « Shear Wave Elastography » (SWE). Afin de comprendre l’origine des modifications observées in vivo, des paramètres standard comme les niveaux de fibrose ou de nécrose ont été caractérisés ex vivo dans le tissu tumoral, grâce à l’immunohistochimie, une technique de référence. / Tumor development is complex process made possible thanks to the microenvironment surrounding tumor cell. Modifications induced by tumor cells on their environment enable their own development by remodeling tissues sustaining them and by creating a new vascular network (angiogenesis). The use of several antiangiogenic therapies, inhibiting the sprout of a new vascular network, has been authorized in clinic. These therapies induce strong modifications in tumors at the functional level and following tumor size changes are is not sufficient to fully characterize tumor modifications. The main goal of this thesis was to use different ultrasound-based imaging modalities in order to assess their sensitivity to modifications induced in murine tumor model (colorectal and lung carcinomas) during different type of therapy (chemical: cytotoxic, antiangiogenic / physical: cold plasma, sonosensitization). Modifications of the spatial distribution of microvessels and their functionality were characterized using contrast-enhanced ultrasound (CEUS), alteration of tumor microstructure was assessed using spectral analysis of radiofrequency signal, known as quantitative ultrasound (QUS) and finally variations of mechanical properties in tumor tissues were measured in shear wave elastography (SWE). In order to better understand the origin of the modifications observed in vivo, standard parameters such as level of fibrosis and necrosis were characterize ex vivo in tumor tissue using immunochemistry as gold standard.

La quantification de la perfusion tissulaire à partir de données dynamiques d’échographie de contraste repose sur une modélisation appropriée de la cinétique de la concentration en agent de contraste dans le tissu étudié. De nombreux facteurs, expérimentaux ou physiologiques, rendent la comparaison inter ou intra-individu de ces paramètres de perfusion difficile. Dans cette thèse, la reproductibilité et la comparaison de différentes méthodes de quantification ont été étudiées dans le cadre d’une étude préclinique de test-retest et sur des simulations numériques. Les méthodes étudiées ont été : le modèle log-normal, le modèle compartimental avec fonction d’entrée et le modèle compartimental avec tissu de référence. Les études précliniques ont montré la difficulté d’estimation d’une fonction d’entrée artérielle et la nécessité de corriger localement le temps d’arrivée de l’agent de contraste dans le tissu pour que l’approximation des cinétiques expérimentales par le modèle soit de qualité suffisante. Une estimation linéaire sous contrainte des paramètres du modèle compartimental avec tissu de référence tirant profit de différentes zones d’intérêt dans l’image a été ensuite proposée pour obtenir à l’échelle régionale et/ou locale des valeurs relatives cohérentes de débit sanguin tissulaire et de volume sanguin tissulaire, exprimées par rapport aux valeurs dans le tissu de référence. Il a été montré que cette approche est la plus robuste et la plus reproductible. L’influence des facteurs tels que la durée d’acquisition, la fréquence d’échantillonnage, le nombre de régions utilisées et l’amplitude du bruit a été étudiée sur des simulations et a permis de formuler des recommandations pour l’acquisition et le traitement des études en échographie de contraste. / Quantification of tissue perfusion from dynamic contrast-enhanced ultrasound data relies on appropriate modeling of the curve representing the evolution of the contrast-agent concentration inside the studied tissue. Many factors, experimental or physiological, make inter-subject or intra-subject comparison of these perfusion parameters difficult. In this thesis, the reproducibility and the comparison of various quantification methods was investigated through preclinical test-retest experiments and through simulations. The investigated methods were: the log-normal model, the one-compartment model using an arterial input function, and the one-compartment model using a reference tissue. The preclinical experiments revealed the difficulty to estimate an arterial input function directly from the image, as well as the necessity to locally correct for the time of arrival of the contrast agent in the tissue in order to ensure the model accurately fits the experimental enhancement curves. A regularized linear estimation of the parameters of the one-compartment model using a reference tissue taking advantage of multiple tissue regions was then proposed to obtain homogeneous relative values of the tissue blood flow and tissue blood volume, expressed relatively to the parameter value inside the reference tissue. The improved robustness and reproducibility of the method was demonstrated. The influence of factors such as the exam duration, the sampling frequency, the number of tissue regions in the analysis, and the noise amplitude were investigated through simulations, and allowed us to formulate recommendations regarding the acquisition and the analysis of contrast-enhanced ultrasound studies.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第22724号 / 医博第4642号 / 新制||医||1045(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 戸井 雅和, 教授 小川 修, 教授 黒田 知宏 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第22735号 / 医博第4653号 / 新制||医||1046(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 羽賀 博典, 教授 伊達 洋至, 教授 万代 昌紀 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 新制・課程博士 / 博士(医学) / 甲第23469号 / 医博第4776号 / 新制||医||1053(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 溝脇 尚志, 教授 武藤 学, 教授 森田 智視 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM