A Patient-Adaptive, High MI Abdominal Scanner We propose to construct and clinically evaluate an adaptive ultrasonic scanner that quickly and automatically adjusts system controls to optimize image quality and assists the sonographer in selecting a favorable acoustic window. We hypothesize that the quality of adaptively-optimized and guided window-selection images will exceed those acquired under conventional scanning conditions. We will test these hypotheses on a modified commercial scanner under the realistic clinical condition of Hepatocellular Carcinoma (HCC) screening. Optimized images will have rapidly- and adaptively-selected transmit power, frequency, focal depth(s), imaging mode (fundamental or harmonic) and other imaging parameters and will be acquired at two Mechanical Indices (the manufacturer’s default setting (MI=1.2), and the “patient-optimized” MI up to a limit of 2.5). On a significant subset of patients, our previous work has shown significant image quality improvements and increased depths of penetration associated with increased MI levels. Our initial studies, presented in this application, show the potential clinical benefits of automated selection of MI and other imaging parameters. Automated selection of MI, as proposed, will realize the ALARA (As Low as Reasonably Achievable) principle for acoustic exposure. Currently, sonographers acquire dozens of individual images during HCC screening for physician review and documentation. A number of published studies and our experience indicate that sonographers use system controls quite sparingly, especially the transmit power level control. Automated selection of imaging parameters and guided selection of acoustic windows should not only improve image quality and depth of penetration, but should also improve the efficiency of scanning procedures and reduce sonographers’ ergonomic challenges. Our initial results and the clinical literature also demonstrate the importance of acoustic window selection in improving image quality and the physical challenges that this task presents to sonographers using current methods, especially in overweight and obese patients. We propose to use the spatial coherence of backscattered echo signals as an image quality feedback parameter. Temporal coherence reflects the electronic SNR and can be used to measure the effective imaging depth in the liver. Our newly developed image quality metric, Lag One Coherence (LOC), quantifies the combined image-degrading effects of reverberation, off-axis scatterers, phase aberration and limited SNR. Our initial phantom and in vivo data demonstrate the robustness of the LOC image quality metric in rapidly determining the optimum patient-specific settings for transmit power, harmonic vs. fundamental imaging, focal depth, and frequency. Our initial data also supports the utility of the LOC in the real-time assessment of the quality of various acoustic windows. We propose to further explore the optimization of these and other imaging parameters and to develop pulse sequences and algorithms to efficiently estimate their preferred settings.

Narrative Currently, operators of ultrasonic scanners must take dozens of images of each patient’s liver and, ideally, adjust the scanner’s controls each time to make the best images. We propose to build and evaluate a system that automatically adjusts most of the controls. We have designed clinical studies in which patients at risk for liver cancer are imaged to see if the automated system can make better images than those that the operator makes.