|
Abstract:
|
Recently, radio-genomics as a new research direction has emerged in order to address the association between imaging data and genomic data in cancer research. Besides, the prediction of clinical outcome is also of interest to many researcher in the field of cancer study. Combining the imaging and genomics data is a better way to assist predicting the clinical outcome. Here we propose a joint multivariate statistical modeling method, which can be applied in the study of the associations among genomic, imaging and clinical outcome. Our finding indicated satisfactory prediction performance, and the joint model provided a rigorous statistical framework to support radio-genomics study.
|
