人工智能在肿瘤病理诊断和评估中的应用与思考

发布时间：2018-05-04 09:15

  本文选题：人工智能 + 病理诊断　； 参考：《第二军医大学学报》2017年11期

【摘要】：人工智能在多个医学场景如疾病诊断、药物筛选、影像医学和护理医学等领域中取得了革命性的进步。病理切片属于二维图像,是人工智能的首要突破点。我国的医疗资源和病理资源丰富,而病理切片的标准化和数字化为人工智能的深度学习提供了大数据背景。我们在乳腺癌、胃癌和胆管癌病理人工智能方面进行了一系列研究,建立了标准的肿瘤细胞标注流程和深度学习流程,研发了肝门部胆管癌人工智能模型,但也发现了存在的问题,提出了解决方案。随着精准性的提高,病理人工智能有望很快进入临床实践。
[Abstract]:Artificial intelligence has made revolutionary progress in many medical settings such as disease diagnosis, drug screening, imaging medicine and nursing medicine. Pathological section belongs to two-dimensional image and is the first breakthrough point of artificial intelligence. China is rich in medical and pathological resources, and the standardization and digitization of pathological sections provide big data background for the deep learning of artificial intelligence. We have conducted a series of studies on the pathological artificial intelligence of breast cancer, gastric cancer and cholangiocarcinoma, established the standard tumor cell labeling process and depth learning process, and developed the artificial intelligence model of hilar cholangiocarcinoma. But also found the existing problems, put forward a solution. With the improvement of accuracy, pathological artificial intelligence is expected to enter clinical practice soon.
【作者单位】： 上海中医药大学附属龙华医院肿瘤科;第二军医大学长海医院病理科;
【基金】：国家自然科学基金(30901794,81572856) 上海市浦江人才计划项目(13PJD002) 上海中医药大学附属龙华医院高层次人才引进项目(LH02.51.002)~~
【分类号】：R730.43;TP18
，

本文编号：1842469