No description available.

No description available.

No description available.

No description available.

No description available.

Prenatal hydronephrosis is a common condition that involves the accumulation of urine with consequent dilatation of the collecting system in fetal infants. There are several hydronephrosis classifications, however, all grading systems suffer from reliability issues as they contain subjective criteria. The severity of hydronephrosis impacts treatment and follow up times and can therefore directly influence a patient’s well-being and quality of care. Considering the importance of accurate diagnosis, it is concerning that no accurate, reliable or objective grading system exists. We believe that developing a convolutional neural network (CNN) based diagnostic aid for hydronephrosis will improve physicians’ objectivity, inter-rater reliability and accuracy. Developing CNN based diagnostic aid for ultrasound images has not been done before. Therefore, the current thesis conducted two studies using a database of 4670 renal ultrasound images to investigate two important methodological considerations: ultrasound image preprocessing and model architecture. We first investigated whether image segmentation and textural extraction are beneficial and improve performance when they are applied to CNN input images. Our results showed that neither preprocessing technique improved performance, and therefore might not be required when using CNN for ultrasound image classification. Our search for an optimal architecture resulted in a model with 49% 5-way classification accuracy. Further investigation revealed that images in our database had been mislabelled, and thus impacted model training and testing. Although our current best model is not ready for use as diagnostic aid, it can be used to verify the accuracy of our labels. Overall, these studies have provided insight into developing a diagnostic aid for hydronephrosis. Once our images and their respective labels have been verified, we can further optimize our model architecture by conducting an exhaustive search. We hypothesize that these two changes will significantly improve model performance and bring our diagnostic aid closer to clinical application. / Thesis / Master of Science (MSc) / Prenatal hydronephrosis is a serious condition that affects the kidneys of fetal infants and is graded using renal ultrasound. The severity of hydronephrosis impacts treatment and follow-up times. However, all grading systems suffer from reliability issues. Improving diagnostic reliability is important for patient well-being. We believe that developing a computer-based diagnostic aid is a promising option to do so. We conducted two studies to investigate how ultrasound images should be processed, and how the algorithm that produces the functionality of the aid should be designed. We found that two common recommendations for ultrasound processing did not improve model performance and therefore need not be applied. Our best performing algorithm had a classification accuracy of 49%. However, we found that several images in our database were mislabelled, which impacted accuracy metrics. Once our images and their labels have been verified, we can further optimize our algorithm’s design to improve its accuracy.

Ultrasound contrast agents have the ability to provide locally increased echogenicity, improving the sensitivity and specificity of images. Due to the unique interaction of microbubbles with the imaging ultrasound field, contrast ultrasonography offers both improved diagnostic techniques, and the potential therapeutic uses of gene and drug delivery through the use of targeted agents. By enhancing the contrast at the tissue-blood interface, an improved image of the structure of organs can be achieved, which is useful in many areas of medical ultrasound imaging. Monitoring the flow of contrast agent in the blood stream also offers information on the degree of blood perfusion into an organ or microvasculature. Present knowledge of the interaction of microbubbles with ultrasound is far from complete. The full potential of contrast agents in improving diagnostic and therapeutic techniques has therefore not yet been achieved. The nonlinear and dynamic properties of microbubble response offer potentially large improvements in contrast to tissue ratio, through intelligent pulse sequence design and/or improved signal processing. Due to various drawbacks of populations studies, only by studying the response from single microbubbles can the interaction be fully understood. The variations of microbubble size and shell parameters within a typical sample of contrast agent dictate that a large number of single scatterer data are necessary to obtain information on the variability of microbubble response, which is not possible with current optical systems. This thesis aims to be a contribution to the understanding of contrast behaviour in response to medical imaging ultrasound pulses. A fully characterized microacoustic system, employing a wide-band piezoelectric transducer from a commercial ultrasound imaging system, is introduced, which enables the measurement of single scattering events. Single microbubble signals from two commercially available contrast agents, Definity R and biSphereTM, have been measured experimentally in response to a range of clinically relevant imaging parameters. The data has been analyzed, together with the results from appropriate theoretical models, in order to gain physical insight into the evolution and dynamics of microbubble signals. A theoretical model for the lipid shelled agent Definity has been developed, and the predicted response from a real sample of single microbubbles investigated. Various characteristics of resonant scatter have been identified, and used to distinguish resonant scatter in experimental acoustic single bubble data for the first time. A clear distinction between the populations of resonant and off-resonant scatter has been observed for a range of incident frequencies and acoustic pressures. Results from consecutive imaging pulses have been used to gain understanding of how initial size, shell material and encapsulated gas may effect the lifetime of a microbubble signal. The response to a basic pulse sequence is also investigated, and an alternative processing method which takes advantage of observed behaviour is presented. Improved understanding of the contrast-ultrasound interaction will provide the basis for improved signal processing tools for contrast enhanced imaging, with potential benefits to both diagnostic techniques and microbubble manufacture.

An adaptive pre-clinical x-ray computed tomography system, named "FaCT" was designed, built, and tested at the University of Arizona's Center for Gamma-Ray Imaging (CGRI). The FaCT system possesses the unique ability to change its magnification and dynamically mask the x-ray beam profile. Using these two abilities, the FaCT system can adapt its configuration to the object being imaged, and the task being performed, while achieving a reduction in the radiation dose applied for imaging.Development of the system included the design of all mechanical components, motion systems, and safety systems. It also included system integration of all electronics, motors, and communication channels. Control software was developed for the system and several high-performance reconstruction algorithms were implemented on graphics processing units for reconstructing tomographic data sets acquired by the system. A new geometrical calibration method was developed for calibrating the system that makes use of the full image data gathered by the system and does not rely on markers.An adaptive imaging procedure consisting of a preliminary scout scan, human guidance, and a diagnostic quality scan was developed for imaging small volumes of interest in the interior of an object at substantially reduced dose. The adaptive imaging procedure makes use of FaCT's adjustable magnification, beam-masking capability, and high-performance reconstruction software to achieve high-quality reconstruction of a volume of interest with less dose than would be required by a traditional x-ray computed tomography system without adaptive capabilities.To address ongoing research into mathematical rules for adapting an imaging system, such as FaCT, to better perform a given estimation task, a method of quantifying a system's ability to estimate a parameter of interest in the presence of nuisance parameters based on the Fisher Information was proposed. The method requires a statistical model of object variability. Possible strategies for increasing the performance of an estimation task, given an adaptive system, were suggested.

No description available.

No description available.