Bibault J-E, Burgun A, Giraud P
Service d'oncologie radiothérapie, hôpital européen Georges-Pompidou, 20, rue Leblanc, 75015 Paris, France; Université Paris Descartes, Paris Sorbonne Cité, 20, rue Leblanc, 75015 Paris, France.
Université Paris Descartes, Paris Sorbonne Cité, 20, rue Leblanc, 75015 Paris, France; Service d'informatique biomédicale, hôpital européen Georges-Pompidou, 20, rue Leblanc, 75015 Paris, France; Inserm, UMR 1138 Team 22 information sciences to support personalized medicine, 20, rue Leblanc, 75015 Paris, France.
Cancer Radiother. 2017 May;21(3):239-243. doi: 10.1016/j.canrad.2016.09.021. Epub 2017 Apr 20.
Performing randomised comparative clinical trials in radiation oncology remains a challenge when new treatment modalities become available. One of the most recent examples is the lack of phase III trials demonstrating the superiority of intensity-modulated radiation therapy in most of its current indications. A new paradigm is developing that consists in the mining of large databases to answer clinical or translational issues. Beyond national databases (such as SEER or NCDB), that often lack the necessary level of details on the population studied or the treatments performed, electronic health records can be used to create detailed phenotypic profiles of any patients. In parallel, the Record-and-Verify Systems used in radiation oncology precisely document the planned and performed treatments. Artificial Intelligence and machine learning algorithms can be used to incrementally analyse these data in order to generate hypothesis to better personalize treatments. This review discusses how these methods have already been used in previous studies.
当新的治疗方式出现时，在放射肿瘤学中开展随机对照临床试验仍然是一项挑战。最近的一个例子是，缺乏III期试验来证明调强放射治疗在其目前大多数适应症中的优越性。一种新的模式正在形成，即通过挖掘大型数据库来回答临床或转化问题。除了国家数据库（如SEER或NCDB），这些数据库往往缺乏所研究人群或所实施治疗的必要详细程度，电子健康记录可用于创建任何患者的详细表型概况。与此同时，放射肿瘤学中使用的记录与验证系统精确记录了计划和实施的治疗。人工智能和机器学习算法可用于逐步分析这些数据，以生成假设，从而更好地实现个性化治疗。本综述讨论了这些方法在以往研究中是如何应用的。
