Improving Liver Ultrasound Image Quality in Difficult-to-Image Patients

提高难以成像患者的肝脏超声图像质量

• 批准号: 10410471
• 负责人: Jeremy Dahl
• 金额: $ 55.13万
• 依托单位: STANFORD UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-08-15 至 2024-04-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10410471
• 关键词: Abdomen; Acoustics; Affect; American; Architecture; Area; Attenuated; Cardiac; Cirrhosis; Clinical; Computer software; Connective Tissue; Data; Development; Diffuse; Disease; Fatty acid glycerol esters; Goals; Healthcare; Healthcare Systems; Heterogeneity; Image; Imaging technology; Lesion; Liver; Liver diseases; Machine Learning; Magnetic Resonance Imaging; Maps; Medical Imaging; Methodology; Methods; Modeling; Noise; Obesity; Output; Overweight; Patient imaging; Patients; Performance; Phase; Physiologic pulse; Pilot Projects; Population; Prevalence; Primary carcinoma of the liver cells; Resolution; Risk Factors; Signal Transduction; Source; Speed; Subcutaneous Tissue; Surveys; System; Techniques; Technology; Testing; Thick; Thyroid Gland; Time; Tissues; Ultrasonography; United States; Weight; clinical imaging; elastography; epidemiology study; fetal; high body mass index; imaging system; improved; in vivo; liver imaging; neural network; novel; obese patients; obese person; patient population; prototype; radio frequency; simulation; sound; subcutaneous; ultrasound

ABSTRACT The prevalence of obesity in the United States has risen to record levels over the past 40 years, putting strain on the healthcare system and creating difficult challenges for medical imaging. We propose to overcome the challenges that obesity poses to ultrasound imaging by (1) developing novel image-quality improvement techniques, and (2) implementing them on pulse-echo ultrasound imaging systems to yield high-quality images of the liver. Ultrasound imaging is uniquely affected by the presence of additional connective tissue and thick subcutaneous fat layers in overweight and obese patients; these additional subcutaneous layers greatly exacerbate reverberation and phase-aberration of the acoustic wave, leading to high levels of clutter, degraded resolution, and overall poor-quality ultrasound images. Our proposed methods will determine the local speed-of-sound in abdominal tissue layers and use this information to accomplish distributed phase-aberration correction. We also apply machine learning techniques to model and suppress the effects of reverberation clutter and speckle noise. The combination of these techniques is expected to achieve significant improvements in liver image quality. These image-quality improvement methods will be implemented on a real-time ultrasound scanner and will be evaluated in clinical imaging tasks of overweight and obese patients undergoing ultrasound surveillance of hepatocellular carcinoma. Successful development of the proposed technology will not only enable high-quality ultrasound imaging of the liver in otherwise difficult-to-image overweight and obese patients, but also facilitate improved image quality across nearly all ultrasound imaging applications, for all populations.

摘要 在过去的40年里，美国肥胖症的流行率已经上升到创纪录的水平， 给医疗保健系统带来压力，并给医学成像带来困难的挑战。我们 建议通过以下方式克服肥胖给超声成像带来的挑战：（1）开发 新的图像质量改善技术，以及（2）在脉冲回波上实现它们 超声成像系统，以产生高质量的肝脏图像。 超声成像是唯一的影响，存在额外的结缔组织和厚 超重和肥胖患者的皮下脂肪层;这些额外的皮下层 大大加剧了声波的混响和相位畸变，导致高 杂波水平、分辨率降低和总体质量差的超声图像。我们提出的 方法将确定腹部组织层中的局部声速， 信息，以完成分布式相位畸变校正。我们也应用机器 学习技术来建模和抑制混响杂波和斑点噪声的影响。 这些技术的结合有望实现肝脏的显着改善 图像质量这些图像质量改善方法将在实时 超声扫描仪，并将在超重和肥胖的临床成像任务进行评估 肝细胞癌超声监测患者。 所提出的技术的成功开发不仅可以实现高质量的超声 在其他方面难以成像的超重和肥胖患者的肝脏成像，而且 在几乎所有的超声成像应用中， 人口。