Molecular and Functional Imaging

Since magnetic resonance imaging (MRI) has a formidable array of capabilities to characterize function and can provide a wealth of spatio-temporal information on the tumor microenvironment, it is well-suited to investigating a complex disease such as cancer. We have developed molecular and functional imaging techniques for revealing key aspects of the tumor microenvironment such as angiogenesis (Kim et al. 2013), lymphangiogenesis (Pathak et al. 2004), the extracellular matrix (Pathak et al. 2005), metastasis (Pathak et al. 2006), and assessing treatment efficacy (e.g. antiangiogenic therapies) in a wide array of preclinical cancer models (Kim et al. 2014). These advances included some of the first MRI methods for imaging transport through the tumor’s extracellular matrix (ECM) (Pathak et al. 2005) and ‘image-based phenotyping’ of tumor microenvironmental remodeling induced by cytokine overexpression (e.g. VEGF) (Pathak et al. 2013). Many of these MRI biomarkers are translatable into the clinic, and compatible with “bench to bedside” applications.

More recently, in collaboration with Dr. Thakor’s Neuroengineering Laboratory, we have employed Laser Speckle Imaging (LSI) to characterize the structural and functional remodeling of the microvasculature during pathological angiogenesis (Rege et al. 2012). We are currently using LSI in preclinical brain tumor and stroke models to better understand the role of the neurovasculature.

Clinical Biomarker Development

If available, non-invasive biomarkers of the hallmarks of cancer (e.g. angiogenesis, hypoxia etc.) could greatly improve clinical decision making, patients’ prognosis and their quality of life. However, the development of non-invasive biomarkers requires a detailed understanding of the complex interplay between cancer biology and the physics of image formation. Therefore, we have developed computational tools for relating MR image contrast to changes in the tumor microenvironment (Pathak et al. 2008).

Image-based Systems Biology

Our growing understanding of cancer pathways has been the result of pioneering studies spanning the cellular, microvascular and tissue levels. However, visualizing these pathways remains a challenge due to the lack of integration between micro- and macroscopic cancer imaging data, which also presents a major hurdle for developing computational models of tumor biology. Therefore, we have initiated the development of ‘multiscale imaging’ for cancer systems biology (Kim et al. 2012). This approach employs imaging methods such as ‘mesoscopic’ scale magnetic resonance microscopy, to bridge the resolution gap between micro- and macroscopic imaging data (Cebulla et al. 2014). Such ‘multiscale’ datasets help correlate the underlying genotype with the emergent tumor phenotype (Kim et al. 2011) and facilitate development of ‘image-based’ models of cancer pathways (Kim et al. 2012). Currently, in collaboration with Dr. Popel’s Systems Biology Laboratory we are using this approach to gain new insight into the role of angiogenesis in breast cancer progression (Stamatelos et al. 2014).

Computational and Visualization Tools

To extract and communicate information from the rich, multidimensional/multimodality/multiscale biomedical images we acquire from preclinical cancer models, it is necessary to harness techniques from information theory, computer graphics, statistics and visualization science. Therefore, we are developing some of the first interactive, image-based atlases of commonly used pre-clinical tumor models. These ‘cloud’ based digital atlases will be made freely available to cancer investigators worldwide, for use in their research applications. Since these atlases can be viewed using freely available software, they can also be used for educational/teaching purposes (Pathak et al. 2011). We are also developing new methods for visualizing preclinical imaging data and numerical tools for exploring the relationship between MRI contrast and the underlying biology.