Using Commercial and Open-Source Tools for Artificial Intelligence: A Case Demonstration on a Complete Radiomics Pipeline

Artificial intelligence (AI) is becoming increasingly used in radiology over the past years due to the astonishing variety of tools and plugins intended for image analysis and modeling. Specifically, the use of radiomics, describing the development of imaging signatures from a number of handcrafted imaging features, is gaining ground since it has shown great promise toward accelerating personalized diagnosis and therapy in healthcare. A de novo synthesis of a radiomics analysis pipeline can be technically complex and requires a carefully designed step-by-step analysis infrastructure, most frequently conducted by AI experts. However, the end result and the benefit of such an analysis can be accessible to an extended number of professionals through the use of user-friendly software tools that do not demand explicit knowledge of the theoretical framework. In view of the above, the aim of this chapter is not only to shed light on the technical and theoretical aspects in radiomics but also primarily to provide the clinical experts with a comprehensive list of user-friendly and readily available tools and plugins to design and experiment on their autonomous end-to-end radiomics analysis workflows.