CRRS-4 Pivotal Somo•v™ ABUS ROC Reader Observer Study

Overview

To determine the impact on Interpreting Physician performance in detecting breast cancer, as defined by the area under the Receiver Operating Characteristic Curve (ROC), when Automated Breast Ultrasound (ABUS) and screening mammography (XRM) are combined, compared to screening mammography alone in asymptomatic women with \>50% parenchymal density and a screening mammogram assigned a BI-RADS Assessment Category 1 (negative) or 2 (normal with benign findings). The effect of the improved Reader performance is illustrated by plotting the ROC curves for XRM alone and XRM+ABUS. If Reader performance improves with the addition of ABUS to the XRM, the area under the curve (AUC) for XRM+ABUS (AUCXRM+ABUS) will be greater than the area under the curve for XRM Alone (AUCXRM Alone). This difference is represented as ∆AUCABUS. The null and alternative hypotheses can be formally expressed as follows: H0: ∆AUCABUS = 0, AUCXRM+ABUS = AUCXRM Alone The null hypothesis is that Reader performance will be unchanged with the addition of ABUS to a screening mammogram assigned a BI-RADS Assessment Category of 1 or 2. HA: ∆AUCABUS ≠ 0, AUCXRM+ABUS ≠ AUCXRM Alone The alternative hypothesis is that Reader performance will be changed with the addition of ABUS to a screening mammogram assigned a BI-RADS Assessment Category of 1 or 2. A statistically significant change will be considered equivalent to a statistically significant improvement if the estimated value of AUCXRM+ABUS is greater than that of AUCXRM Alone with statistical significance at an alpha level of .05 for a two-sided test.

The Clinical Retrospective Reader Study is an observational case-controlled, multi-reader multi-case Receiver Operating Characteristic (ROC) study involving approximately 16 Interpreting Physicians (Readers) and approximately 200 breast screening cases which will be reviewed and of those, 164 breast screening cases will be included in the primary data analysis. All primary data analysis cases will consist of screening mammogram and Automated Breast Ultrasound image readings from the same asymptomatic females with \>50% parenchymal density for whom the screening mammogram was assigned a BI-RADS Assessment Category 1 (negative) or 2 (normal with benign findings). The Readers will review a total of 200 cases: the Non-Cancer Case Set, the Cancer Case Set, and additional Control and Supplemental Case Sets. These cases will be randomized during the reading session so each Reader will review the entire Case Set in a unique order. The primary Receiver Operating Characteristic analysis will be performed on the Readers' Likelihood of Malignancy (LOM) ratings from their initial review of the screening mammogram alone and their final review of the screening mammogram and Automated Breast Ultrasound for the Non-Cancer Case Set and Cancer Case Set.