The proposed portfolio training program consists of four related classes in BME, ECE, and CS that prepare graduates to pursue a research experience in imaging science at UT Austin. Additional elective practical classes have also been identified. Students completing this program might choose to work on a dissertation research problem in either basic or clinical science at The University of Texas at Austin (UT Austin), The University of Texas Health Science Center at Houston (UTHSCH), The University of Texas M. D. Anderson Cancer Center (UTMDACC), The University of Texas Medical Branch at Galveston (UTMB), or the USAF HEDO Research Laboratory at Brooks City-Base (USAF-BCB) in San Antonio, Texas.
The Third Biomedical Imaging Research Opportunities Workshop identified four areas of imaging that present opportunities for research and development: 1) Multimodality Image-Guided Therapy, 2) Imaging Informatics, 3) Imaging Cell Trafficking, and 4) Technology Improvement and Commercialization [1]. The proposed portfolio program addresses all of these areas. Multimodality image-guided therapy is being developed by faculty in the Division of Diagnostic Imaging at UTMDACC for breast and prostate cancer. Imaging Informatics is being developed by faculty at UT Austin, the School of Health Information Science (SHIS) at UTHSCH, and the Department of Biostatistics and Applied Math at UTMDACC. Faculties at all of the participating institutions are developing selective contrast agents and functional 3D vs. time measurements at high spatial resolution to image cell trafficking. Many of the faculty participating in the proposed training program are developing new imaging instrumentation, contrast agents, and models to measure in vivo cellular signaling and enzyme kinetics, and are working with industrial partners to commercialize the new technology. This training program stresses professional development so our trainees can fill critical niches in academic and industrial imaging research.
We envision a comprehensive training program that educates students about the complete life cycle of imaging science, from image formation to analysis and decision-making. In this cycle, one must select a clinically relevant problem and understand the biophysical process to be imaged. Given suitable problem selection, there are four key skill/knowledge sets necessary for designing an imaging solution. First, one must be familiar with available instrumentation and be capable of developing new instrumentation to sense the specified process in the object to be imaged. Second, one must understand the principles of digital signal processing in order to effectively and efficiently extract relevant information from imaging data. Third, one must develop and apply modeling and visualization techniques for interpretation that quantitatively depict “how things work.” Fourth, informatics methods for optimizing the storage and use of biomedical imaging data are needed to support both scientific discovery and clinical decision-making.
A flagship feature of this imaging program is the emphasis placed on computational analysis of imaging data. Ultimately, the only way to predict behavior of complex biological systems is to have accurate mathematical models of those systems. We observe selected anatomical and functional features of the systems using imaging tools. However, there is a complex interaction between the observed features and the underlying physiological processes that produce those features. An imaging scientist must be able to understand the limitations of a data set and how those limitations affect the validity of the models developed. Moreover, this process is highly iterative since the model often suggests the quantifiable features that must be observed, and the model class depends on an understanding of the physiology and the imaging instrumentation. Models and data acquisition must be coupled through multiple scales and organizational levels to integrate molecular, cellular, and organ structure and function with human disease [2-4]. Effective treatment relies on both accurate interpretation of imaging data and the validity of the model of the physiological process being studied.