Laurence (Larry) Clarke, Ph.D.
Branch Chief
Imaging Technology Development Branch
Cancer Imaging Program
EPN, Room 6066
6130 Executive Blvd.,
Rockville, MD 20852-7412
BIP WebPage: New Initiatives: http://imaging.cancer.gov/

Bio:

Dr. Clarke as of January 1999 is the Branch Chief for Imaging Technology Development for the Cancer Imaging Program, Division of Cancer Treatment and Diagnosis, NCI. In this capacity he is responsible for the development of initiatives for new and emerging imaging technology and methodologies as applied to cancer. Dr Clarke is also on a detail assignment at NIBIB and is a guest scientist at NIST. His responsibilities at NCI, NIBIB and NIST include initiatives that support public research resources for assessing new imaging methods including the evaluation of image processing algorithms and methods for the measurement of drug response in particular.

Before joining NCI, Dr Clarke was a Professor of Radiology, with adjunct appointments in computer science and physics at the University of South Florida (USF), and was a full member and imaging program leader at the H. Lee Moffitt Cancer and Research Center at USF. He has previously worked at other cancers centers including University of Miami and the Memorial Sloan Kettering Cancer Center, NYC. Dr. Clarke has been active over the last 30 years in medical imaging technology development, including image processing for early cancer detection, cancer diagnosis and treatment using a range of imaging modalities. He is a Fellow of the ISMRM and AAPM. He graduated with a PhD in physics at the National University of Ireland (1978) and an MSc degree from Queens University of Belfast, Ireland (1968).

Talk Title: Importance of reference image databases to evaluate performance of clinical decision tools

Abstract: This presentation will provide an overview of NCI efforts to create reference image databases as a web accessible public resource to evaluate the performance of clinical decision tools. The long term goal of this effort is to create reference standards that will permit the relative performance of software tools to be evaluated in a standardized manner. A review of the complexity of creating this resource will be examined, such as: (a) harmonization of data collections across imaging platforms to create this resource, (b) the inclusion and exclusion criteria as related to imaging protocols and clinical case selection, (c) methods for image mark up or “truth files” that software results are compared against, and (d) how to objectively train and test software tools against the database. Finally in the context of 3D imaging or 3D segmentation, methods to develop a resource to optimize specific components of software tools used in the clinical decision process will be discussed. Examples of NCI image database that are underway will be presented including the potential role of NIST in these efforts.