R.J. Gillies
H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL USA 33612
I have always been convinced that “Images are Data”. Early in my career I remember being frustrated that entire tumors were being described by a single variable (such as the SUV in PET or the Ktrans in DCE MRI), when it was clear to anyone looking at the images that tumors were highly heterogeneous. Further, as a cancer biologists, I knew that his heterogeneity was a critical factor in therapy response. In 2008, I moved my lab to the Moffitt Cancer Center, who supported me to establish a program in clinical image analytics, which we termed “Radiomics”. At that time, I was also a visiting TEFAF professor at the University of Maastricht, and spent the summer of 2008 in the lab of Dr. Philippe Lambin. During this time, we came up with the outlines of “Radiomics” and the radiomics pipeline, which led to our first annual retreat in 2009, and a subsequent grant from the Quantitative Imaging Network (QIN) at the National Cancer Institute of the NIH. At that time, we could not have foreseen how explosive a discipline “Radiomics” would become. Because of the input of many investigators around the world, the field has matured significantly in the last decade. (…) We are now at an important transition in radiomics wherein the pipelines for processing radiomic data are no longer a major challenge. Thus, we are now faced with developing and validating models that can be clinically impactful, which has its own set of challenges. An advantage provided by Radiomics is that the images (data) are routinely obtained through standard of care. Thus, there is the potential to accrue images (data) from thousands of patients, yet this has its own challenges in data sharing or distributed learning, in which Chinese scientists are uniquely prepared to meet. A further challenge to Radiomics is in providing an underlying biological or pathophysiological rationale to describe and interpret the radiomic image features. When these challenges can be met, the impact of radiomics in clinical care will be fully realized, to the benefit of practitioners, payers, and most importantly to the benefit of patients.
Reference: Extract of the preface of the book 《FOUNDATIONS of RADIOMICS》2020 edited by Dr. Jie Tian, Director of the Key Laboratory of Molecular Imaging, Institute of Automation, Chinese Academy of Sciences,