国际团队最近在《英国杂志》中发表了一项角色，指出他们开发的新人工智能工具可用于预测未来20年中许多疾病的风险，这将帮助医生识别高风险群体并尽早采取预防步骤。 Researchers from the German Cancer Research Center and other institutions have designed an artificial intelligence tool called Delphi-2M, using the health data of 400,000 people in the UK Biological Sample Library to train it, so that it can predict the possibility of a person's medical history suffering from various factors such as age, gender, body mass, body mass and smoking and smoking and smoking and smoking, and smoking and smoking, and smoking and smoking, and smoking and smoking, and smoking and smoking和吸烟。有超过一千种不可预测的疾病。研究人员是 - 启用它也来自丹麦190万人的游戏数据。比较现有AI工具的Delphi-200m表明，其预言精度比目前用于评估单一疾病风险的各种AI工具相似或更好，并且比使用生物标志物来预测许多疾病风险的机械研究算法要好。但是，研究人员还说，该AI工具具有局限性，例如用于培训的数据资源范围狭窄。他们计划从更多国家 /地区进一步训练它，以扩大应用程序的范围。