FIGURE 5.

Individual tool results. ROSE concordance (A) and true positive rates (B) for individual tools. By design, all lesions included in this analysis were ultimately proven to be positive for malignancy based on final pathology results of the ENB-aided sample. Any individual tool yielding only benign or inconclusive results was considered a false negative while any tool yielding at least one malignant result was considered a true positive. C, Individual impact of each individual tool on multimodality success. Among all 416 subjects with single lesions ultimately proven to be true positive for malignancy, this analysis examines the impact of “ignoring” each tool in turn within the analysis. For example, if only the biopsy forceps yielded a malignant result and all other tools yielded negative results, the overall result for that case would be considered negative when the biopsy tool was ignored in the analysis. The impact of ignoring each tool in turn is shown. For example, within the context of the multimodality sampling strategy, if the biopsy forceps had not been used, 9.3% of true positive malignant cases would have been missed (or in other words, adding biopsy forceps to the tool strategy increased the true positive rate by 9.3%). ROSE indicates rapid on-site evaluation.