Center for Innovation in Medicine, Bucharest, Romania.
KOL Medical Media, Bucharest, Romania.
Eur J Public Health. 2022 Jun 1;32(3):443-449. doi: 10.1093/eurpub/ckac016.
Although current efforts are made to diminish the incidence and burden of disease, cancer is still widely identified late at stage. This study aims to conduct a systematic review mapping the existent and emerging clinical research on artificial intelligence (AI) in the treatment of cancer and to underpin its integration challenges and opportunities in the European Union (EU) health sector.
A systematic literature review (SLR) evaluating global clinical trials (CTs; published between 2010 and 2020 or forthcoming) was concluded. Additionally, a horizon scanning (HS) exercise focusing on emerging trends (published between 2017 and 2020) was conducted.
Forty-four CTs were identified and analyzed. Selected CTs were divided into three research areas: (i) potential of AI combined with imaging techniques, (ii) AI's applicability in robotic surgery interventions and (iii) AI's potential in clinical decision making. Twenty-one studies presented an interventional nature, nine papers were observational and 14 articles did not explicitly mention the type of study performed. The papers presented an increased heterogeneity in sample size, type of tumour, type of study and reporting of results. In addition, a shift in research is observed and only a small fraction of studies were completed in the EU. These findings could be further linked to the current socio-economic, political, scientific, technological and environmental state of the EU in regard to AI innovation.
To overcome the challenges threatening the EU's integration of such technology in the healthcare field, new strategies taking into account the EU's socio-economic and political environment are deemed necessary.
尽管目前正在努力降低疾病的发病率和负担,但癌症仍被广泛发现处于晚期。本研究旨在对人工智能(AI)在癌症治疗方面的现有和新兴临床研究进行系统综述,并为欧盟(EU)卫生部门的 AI 整合挑战和机遇提供依据。
进行了一项系统文献综述(SLR),评估了全球临床试验(CT;发表于 2010 年至 2020 年或即将发表)。此外,还进行了一项关注新兴趋势的地平线扫描(HS)研究(发表于 2017 年至 2020 年)。
确定并分析了 44 项 CT。选定的 CT 分为三个研究领域:(i)AI 与成像技术结合的潜力,(ii)AI 在机器人手术干预中的适用性,(iii)AI 在临床决策中的潜力。21 项研究具有干预性质,9 篇论文为观察性研究,14 篇文章未明确提及所进行的研究类型。这些论文在样本量、肿瘤类型、研究类型和结果报告方面表现出较大的异质性。此外,研究方向也发生了转变,只有一小部分研究在欧盟完成。这些发现可能与欧盟在人工智能创新方面当前的社会经济、政治、科学、技术和环境状况进一步相关。
为了克服威胁欧盟将此类技术融入医疗保健领域的挑战,需要考虑到欧盟社会经济和政治环境的新战略。