Bridging the gap between human beings and digital twins in radiology

Radiology is a multimodal discipline that makes extensive use of different imaging methods, producing a huge amount of quantitative, longitudinal, and digital data over time. Radiologists must, therefore, be at the center of the “technological revolution” of healthcare. Modern computational models are boosting the capability to diagnose and prognosticate disease, which will drive individualized treatment plans. This might be achieved by accurately modeling an individual’s progression along a treatment pathway or algorithm—from pretreatment patient selection to customizing treatment nuances and through reevaluation and follow-up [1]. However, presently, the current applications of artificial intelligence (AI) in medicine can only address narrowly defined tasks; this is in contrast to the human ability to think clinically, analyzing and synthesizing data from multiple sources and modalities [2]. In this landscape, the technology that will likely support clinicians in disease prognosis and prediction of response to therapy in the near future are Digital Twin (DT) technologies, also referred to as Virtual Human Twins in Europe.