Biomedical Imaging

BRP staff provide statistical expertise to the Biomedical Imaging Program regarding the design and analysis of trials relating to molecular, diagnostic, and therapeutic imaging. Such trials include technology evaluation and reader variability studies, large screening trials, early-phase studies of novel imaging-contrast agents, and image-guided therapy trials. Staff represent the NCI in the interim monitoring of these trials, as necessary. Additionally, staff collaborate in the design of databases for the development and evaluation of computer-assisted diagnostic (CAD) methods, which is a critical step in obtaining acceptance of such tools in the medical community. BRP staff also conduct research on statistical methods relevant to the imaging community. Current research topics include Receiver-Operator Characteristic (ROC) Curve regression and methods for estimating diagnostic accuracy without a gold standard for establishing true disease status.