The visualization of microcalcifications with medical ultrasound

One of the fundamental limitations of current medical ultrasound technology in the imaging of the breast is its inability to reliably visualize microcalcifications. Improving microcalcification detection with ultrasound would have a significant impact on the management of breast cancer. This dissertation is based on the hypothesis that the visualization of microcalcifications with medical ultrasound is currently limited by a number of factors, including speckle noise, the imaging system spatial resolution, and phase aberration. Investigation of the relative impact of these factors will guide system design optimization. In the interest of this objective, it is hypothesized that an ultrasound imaging system with adequate spatial resolution, phase aberration correction, and speckle reduction capabilities would allow the reliable detection of microcalcifications.; An approach for comparing proposed imaging systems and measuring the impact of phase aberration and tissue backscatter is developed based on detection theory. Initial results indicate that fully coherent, low f-number arrays offer the best performance, and that moderate simulated phase aberration can significantly reduce visualization performance.; The classical Faran model is used to study the acoustic parameters of microcalcifications represented as calcium hydroxyapatite suspended in a fluid. Clinical measurements of echoes from suspected in vivo microcalcifications are presented and compare favorably with model predictions. The spatial and temporal coherence of microcalcification echoes is analyzed and discussed in the context of the impact of spatial and frequency compounding. It is shown through theory and experiment that these speckle reduction techniques only marginally improve microcalcification visualization.; The impact of phase aberration on microcalcification visualization is examined. Two types of aberration are described and studied. These include phase errors arising from both a gross over- or under-estimation of the mean tissue velocity and those arising from tissue inhomogeneities. Both are shown to have considerable impact on visualization. However, aberration due to gross velocity error was found to be the more significant of the two in the in vivo data collected, and is one of the two predominant factors limiting visualization among those studied.; Clinical tissue backscatter measurements are presented and integrated into a comprehensive analysis of the impact of speckle noise and array geometry. This analysis supports the clinical experience, demonstrating that speckle noise is the other of two predominant factors limiting visualization among those studied. Results are presented comparing a typical clinical array to proposed high-performance arrays. These results indicate that aggressively designed two-dimensional arrays would significantly improve the reliability of microcalcification visualization.