Lahmi Lucien, Bibault Jean-Emmanuel, Constantinides Yannis, Azria Joseph, Cheval Véronique, Dejean Catherine, Durdux Catherine, Ducteil Angélique, Escande Alexandre, Gesbert Cédric, Haaser Thibaud, Kobeissi Gabriel, Lemanski Claire, Nataf Moshe, Raynaud Charles, Thureau Sébastien, Lagrange Jean-Léon, Huguet Florence
Service d'oncologie radiothérapie, CRTT, Versailles, France.
Service d'oncologie radiothérapie, hôpital européen Georges-Pompidou, Paris, France.
Cancer Radiother. 2025 Sep-Oct;29(5-6):104673. doi: 10.1016/j.canrad.2025.104673. Epub 2025 Jul 24.
The growing integration of artificial intelligence is profoundly transforming radiation oncology, particularly through the automation of tasks such as target volume delineation. While artificial intelligence holds the promise of enhancing treatment efficiency and accuracy, it also raises significant ethical and educational concerns: How can we preserve core clinical skills? How can we transmit expertise that is now less frequently practiced? How can we ensure truly informed patient consent in the face of automated systems? And finally, how should we use the time freed up by automation? To address these challenges, medical training must increasingly incorporate critical thinking and algorithmic literacy. The caregiver-patient relationship must remain central, with medical decisions remaining under human responsibility. Artificial intelligence should thus remain a tool in the hands of physicians, not a substitute for clinical judgment. Ethical reflection must accompany every step of artificial intelligence implementation, ensuring its integration is thoughtful, responsible, and human-centred.
人工智能日益广泛的整合正在深刻改变放射肿瘤学，特别是通过靶区勾画等任务的自动化。虽然人工智能有望提高治疗效率和准确性，但它也引发了重大的伦理和教育问题：我们如何保留核心临床技能？我们如何传授现在较少实践的专业知识？面对自动化系统，我们如何确保患者真正知情同意？最后，我们应该如何利用自动化节省的时间？为应对这些挑战，医学培训必须越来越多地纳入批判性思维和算法素养。医患关系必须始终处于核心地位，医疗决策仍由人类负责。因此，人工智能应始终是医生手中的工具，而不是临床判断的替代品。在人工智能实施的每一步都必须伴随伦理思考，确保其整合是深思熟虑、负责任且以人为本的。
