Susan G Komen  
I've Been Diagnosed With Breast Cancer Someone I Know Was Diagnosed Share Your Story Join Us And Stay Informed Donate To End Breast Cancer
    Home > Research & Grants > Grants Program > Research Grants > Research Grants Awarded > Abstract

    Research Grants Awarded

    A Comparison Of Diagnostic Performance Of Digital Breast Tomosynthesis (Dbt), Cone Beam Ct (Cbct) And Mri In Women With Known Breast Abnormalities

    Grant Mechanism:
    Investigator Initiated Research

    Scientific Abstract:
    Periodic mammography has been proven as a cost-effective method to detect breast cancer at an early stage and, when followed by an appropriate diagnosis and disease management / treatment to reduce mortality of patients. However, the sensitivity and specificity of conventional mammography are relatively low, in particular for younger women with dense breasts. Based on recent studies, the American Cancer Society (ACS) established guidelines for breast screening of high risk women using breast magnetic resonance imaging (MRI) as an adjunct to the mammography. Recently, several new and innovative breast imaging technologies have been developed and investigated in the hope of improving both detection and diagnostic performance. These include full-field digital mammography (FFDM), whole breast ultrasonography (US), breast digital tomosynthesis (BDT), and breast cone-beam computed tomography (CBCT). Among these, BDT and CBCT generate 3D images similar to MRI. MRI has several disadvantages for this purpose: (1) low or variable specificity in diagnosis of ductal carcinoma in situ (DCIS) and other types of cancers such as lobular, (2) the need for injecting contrast agent that may be harmful to a subset of women and (3) the high professional resource and operational cost. As a result, BDT and CBCT examination may prove to be reasonable alternatives for the detection and diagnosis of breast cancers in particular for but not limited to high risk women, with potentially high spatial resolution at a lower resource utilization and operational cost. Screening MRI in high risk women may offer the advantage of detecting breast cancers that may not be seen in dense breasts. Screening ultrasound has also been shown to detect fewer occult cancers in this population, but ultrasound has limitations created by operator variability. Tomosynthesis and CBCT offer the potential to identify and correctly characterize occult cancers. Little to no data comparing the clinical utility and performance of BDT or CBCT to MRI in either the screening or the diagnostic arenas are available. Better understanding of these technologies in terms of their diagnostic ability and ultimately, their optimal role in the clinical practice is of great importance and interest. We propose to investigate whether using BDT or CBCT can achieve comparable or improved performance as compared to MRI technology in (1) classification of known breast abnormalities into malignant and benign categories and (2) measurements of abnormality sizes (i.e. increasing agreement level between the results generated by each of these imaging modalities and pathology). Specifically, we propose to conduct a comprehensive multi-case, multi-reader, and multi-mode observer performance and observer abnormality size measurement study. APPROACH: We have established large, diverse and carefully verified image databases in our research laboratory. Our clinical breast imaging center is equipped with a large number of full-field digital mammography (FFDM) systems (11), whole breast ultrasonography (US), 2 breast digital tomosynthesis (BDT) systems, 2 breast magnetic resonance imaging (MRI) systems used primarily in breast imaging, and very soon breast cone beam computed tomography (CBCT) system. Our medical center performs over 45,000 screening mammography examinations annually and an additional 40,000 diagnostic imaging procedures due to the fact that we are the largest referral center in the region. Through screening we detect over 150 new cancers each year and in the referral population we see over 800 additional breast cancers each year. Because we serves as a primary referral site for the majority of diagnostic examinations and follow-up (including biopsy if needed) in the region, we perform over 7,000 biopsies per year and we manage approximately 1000 new breast cancer patients per year. We will perform the following tasks: 1. Identify and recruit 200 consenting participant women who underwent standard mammography examinations during which suspicious breast abnormalities were detected and rated as BIRADS 4 or 5. Before undergoing a biopsy, the woman will also undergo bilateral MRI, BDT and CBCT examinations. Images and subsequent pathology assessments obtained from a percutaneous core and/or surgical biopsy will be collected and assembled in a database. 2. Conduct a fully crossed, mode counter balanced observer performance and an abnormality size measurement accuracy study. Eight radiologists will independently read each case three times using MRI, BDT, and CBCT images. Observers will rate the likelihood of malignancy using both BIRADS ratings and a probability scale (0 to 100%) and measure abnormality size (three longest cross-section lengths). 3. Perform MRMC ROC-type analyses to assess differences, if any, in diagnostic performance between using these imaging technologies. We will account for variability and correlations amongst and between readers, modalities and cases. We will also compare accuracy of size measurements when using each of these technologies versus the ?gold-standard?.

    Lay Abstract:
    Periodic mammography has been proven as a cost-effective method to detect breast cancer at an early stage and, when followed by an appropriate diagnosis and disease management / treatment to reduce mortality of patients. However, the sensitivity and specificity of conventional mammography are relatively low, in particular for younger women with dense breasts. Based on recent studies, the American Cancer Society (ACS) established guidelines for breast screening of high risk women using breast magnetic resonance imaging (MRI) as an adjunct to mammography. Despite it?s highly publicized performance, MRI has several disadvantages for this purpose: (1) low or variable ability to diagnose some cancers associated with micro-calcifications and other types of cancers, (2) the need for injecting contrast agent that may be harmful to a subset of women (3) the high professional resource and operational cost and (4) limited access in some regions. Recently, several new and innovative breast imaging technologies have been developed and investigated in the hope of improving both detection and diagnostic performance. Two of these, namely, breast digital tomosynthesis (BDT), and breast cone-beam computed tomography (CBCT) generate 3D images similar to MRI. . Little to no data comparing the clinical utility and performance of BDT or CBCT to MRI in either the screening or the diagnostic arenas are available. Better understanding of these technologies in terms of their diagnostic ability and ultimately, their optimal role in the clinical practice is of great importance and interest. We propose to investigate whether using BDT or CBCT can achieve comparable or improved performance as compared to MRI. The way this will be done is by collecting a large number of examinations with known diagnosis and then performing a study in which radiologists read these cases independently without knowing the ?truth?. Analysis of their performance allows for such a comparison to be made. Therefore the study includes a number of tasks including; recruiting women participants, performing multiple examinations on those who agree to participant, assembling a comprehensive dataset of examinations, running an observer (radiologist) performance study to assess their accuracy when using these examinations and analyzing and comparing the results. If any of these technologies (or perhaps a combination of these) can achieve comparable or better performance than MRI alone, in either classification of the suspected abnormality as likely to be benign or malignant or in measuring its size, then these technologies may ultimately be considered as viable alternatives to MRI imaging in large subsets of the women. It is conceivable that one or both of these tests could substantially lower operational costs and complexity while simultaneously, improve access and maintain and perhaps even improve quality of care. This is, to our knowledge, the first of these studies and therefore should be viewed as a preliminary yet an important and warranted study at this stage in the development and evaluation of these important and very advanced imaging technologies.