AI in Radiotherapy
摘要
Radiation therapy (RT), or radiotherapy, is an integral and uniquely technologically reliant cornerstone of cancer care. AI is poised to transform many aspects of the RT patient trajectory. In each step, there is an abundance of patient data ranging from demographic, imaging, clinical, temporal, and dosimetric data. The scope of existing and emerging applications working to harness this data ranges from radiomics, clinical decision support systems, dose prediction and adjustments, predictive modelling and outcome predictions, treatment planning and optimisation, automated contouring and segmentation, adaptive radiation therapy, and quality assurance. While many applications for RT care overlap with medical imaging solutions reported in other chapters, there are other diverse applications that can serve the RT patient and professional context uniquely. This chapter covers the value of standardisation of radiotherapy big data and the aligned value to re-irradiation, emergency preparedness, and other system- and patient-level considerations. It then outlines four broad applications for AI in RT: the leveraging of medical imaging AI solutions (including radiomics and synthetic CT), treatment planning (including auto-segmentation and auto-planning), optimisation of operations, and patient reported outcomes. Given the critical importance of collaborative expertise in the implementation of AI, the chapter also includes an overview of how the RT team can effectively approach realignment of workflows and roles to ensure that AI is implemented in the best interests of the patient. AI is increasingly exhibiting impact across the RT care trajectory and within the involved professions. Strategies developed and leveraged in other areas of healthcare and broader society, including system-level harnessing of big data, radiomics, automation, and predictive analytics, are all poised to transform the ability to deliver high-quality, patient-centered precision medicine. This chapter is designed to provide a high-level picture of the various areas in which AI is being explored and implemented, advancing the science of RT technology and care.