3D Multimodal Ultrasound Imaging for Breast Cancer Screening in Women with Dense Breasts
Abstract
Early detection remains one of our best defenses against breast cancer. A woman’s 5-year survival rate is 99% if the cancer is detected in the early stage, as opposed to 27% if found in late stage. Currently, the most important screening test for breast cancer is mammography (X-ray imaging of breasts). Annual screening mammograms are recommended for women starting at the age of 40. Mammography, however, is not a perfect test. It has a high false positive rate (~50%) and it is not sensitive in patients with dense breasts. In the United States, more than 40% of women have dense breasts. Dense breast tissue not only masks cancerous tumors in mammograms, but also indicates increased risk of breast cancer. However, mammography misses 4 out of 7 cancers per 1000 women screened in this population. Motivated by the clinical evidence and elevated awareness of cancer risks associated with dense breasts, 38 US states and the District of Columbia have passed breast density notification laws. These laws require healthcare providers to inform women with dense breasts of the additional screening or tests that they should consider. These tests typically include three-dimensional (3D) digital mammography, magnetic resonance imaging (MRI), ultrasound, or molecular breast imaging (MBI). Unfortunately, although the added tests do detect more cancers, they also produce a large number of false positives: more than 80% suspicious findings turn out to be benign. As a result, a large number of patients undergo unnecessary follow-up tests and biopsies that could have been avoided. Coping with the fear of having cancer and going through biopsies can frequently cause emotional and physical distress, sometimes long-lasting. These unnecessary tests and procedures also incur hundreds of millions of dollars in costs each year. To conquer the problem of overdiagnosis and the consequent overtreatment of breast cancer, here we propose to develop a new ultrasound breast cancer screening technique that can be used together with mammograms to detect cancer more accurately. Ultrasound has been shown clinically to be sensitive of detecting small and mammographically occult cancers. It has great potential to benefit a large population because ultrasound is low-cost, widely available, well-tolerated by patients, and does not involve ionizing radiation or intravenous injection of contrast. However, ultrasound is not currently used in routine clinical screening because traditional grayscale ultrasound imaging suffers from the same issue of high false positives as other imaging modalities. Meanwhile, clinical ultrasound is mostly done with a hand-held transducer and the images are only two-dimensional (2D); therefore, the experience and competence of the operator (i.e., sonographer or physician) can sometimes influence the imaging results. Over the last two decades, many new ultrasound imaging techniques have been developed to address the issues of high false positives and operator dependence. Among these methods, it was shown that when combining grayscale images with tissue stiffness images (obtained using a technique called ultrasound elastography) and/or tissue microvascular images (obtained using a technique called ultrasensitive microvessel imaging), specificity of ultrasound is significantly improved. It was also found that 3D imaging was able to mitigate the issue of operator dependence for ultrasound. Based on these developments and findings, in this project we plan to develop a new multimodal imaging technology that: 1) combines the effective imaging modes for breast cancer detection including grayscale imaging, elastography, and microvessel imaging; 2) enhances the clinical performance of all imaging modes by realizing 3D imaging for comprehensive breast lesion evaluation; and 3) develops a new machine learning-based image analysis method to classify breast lesions with higher accuracy. Our hypothesis is that by combining the effective ultrasound bioma
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Dec 05, 2021
- Source ID
- W81XWH2110063
Entities
People
- Shigao Chen
Organizations
- Mayo Clinic
- United States Army