OPTIMISING COLORECTAL CANCER PREVENTION TROUGH PERSONALISED TREATMENT WITH ARTIFICIAL INTELLIGENCE

人工智能（AI）技术与临床医学的整合一直受到与不确定的长期临床效益、成本效益和伦理/法律问题相关的挑战的阻碍。人工智能在结肠镜检查中的应用作为一个前沿研究课题，也一直受到这些问题的困扰。然而，人工智能有望通过个性化治疗降低结直肠癌的发病率和死亡率。我们提出的项目将改变这种令人沮丧的状况，从而大大改善结直肠癌的治疗。我们实现了以下目标：通过开展一项有222,000名参与者的泛欧人群随机试验，确定人工智能辅助结肠镜检查在结直肠癌预防中的价值。2. 开发用于结肠直肠息肉和癌症个性化治疗的结肠镜人工智能风险预测工具。3. 开发人工智能辅助结肠镜在结直肠癌筛查中的成本效益模型。4. 调查人工智能开发和实施中的道德和法律障碍。5. 生成第一个可信赖且快速更新的(？人工智能筛查结肠镜检查的临床指南。6. 通过将最终用户纳入人工智能开发和沟通过程，开发面向患者的人工智能工具。我们的研究联盟最大限度地提高了项目的可行性，来自9个欧洲国家、美国和日本的世界级专家在胃肠病学、机器学习、癌症筛查、生物统计学、疾病建模、监管科学、医学伦理和法律方面。我们还与欧洲一家中型人工智能企业合作，实施项目产出。该项目对欧洲乃至世界的患者、社会和经济都有巨大的潜在影响。一旦这项技术在欧洲被广泛接受，预计每年可减少6000人死于结直肠癌，节省7.2亿欧元

ntegration of artificial intelligence (AI) technologies into clinical medicine has been hampered by challenges related to uncertain long term clinical benefits, cost-effectiveness, and ethics/legal concerns. Application of AI in colonoscopy, a forefront research topic, has been also bothered by these issues. AI is, however, expected to reduce colorectal cancer incidence and mortality through personalised treatment. Our proposed project would change this discouraging situation and thus drastically improves colorectal cancer care. We achieve the following goals: 1. Establish the value of AI-assisted colonoscopy in colorectal cancer prevention by conducting a pan-European, population-based, randomised trial with 222,000 participants. 2. Develop a colonoscopy AI risk-prediction tool for personalized treatment of colorectal polyps and cancer. 3. Develop cost-effectiveness models of AI-assisted colonoscopy in colorectal cancer screening. 4. Investigate ethical and legal barriers in AI development and implementation. 5. Generate the first trustworthy and rapidly updating (? living?) clinical guidelines for AI in screening colonoscopy. 6. Develop a patient-oriented AI tool by including end-users in the process of AI development and communication. Our research consortium maximises the feasibility of the project, with world-class experts in gastroenterology, machine learning, cancer screening, biostatistics, disease modelling, regulatory science, and ethics and law in medicine from nine European countries, the US and Japan. We also partner with a European medium-size AI enterprise for implementing the output of the project. The project has large potential impact on patients, society, and economy in Europe, and the World. Up to 6,000 fewer deaths due to colorectal cancer and saving of 720 million euro can be expected per year once the technology is widely accepted in Europe