Research
Multi-resolution wavelet kinetic features used to characterize intratumor heterogeneity
Research Overview
Work in our group focuses on developing innovative image analysis, machine learning and data science methodologies for multimodality imaging, and also on incorporating such methods into clinically relevant applications. To this end, we have developed innovative computational methodologies that have enabled the investigation of novel phenotypic biomarkers via imaging, as well as translating these biomarkers through large clinical and epidemiological studies, to address relevant research questions for personalizing cancer care.
Specific thematic areas of research include the emerging fields of radiomics and radiogenomics, investigating both the association of such phenotypic imaging markers with genetic markers well as their potential to augment current standard clinical assessment, and most importantly emerging molecular and histopathologic biomarkers in breast cancer risk assessment, determining prognosis, and predicting patient therapy response and survival.
Most of our work to date has been on breast and lung imaging, with recent expansion into other oncologic imaging areas, including ovarian, pancreatic, and skin cancer. Molecular imaging applications is also an emerging area of interest.