蜜蜂:通过基础模型驱动的嵌入,在瘤学中实现可扩展的多式联络AI
Aakash Tripathi1,2, Asim Waqas3,4, Matthew B Schabath4
1Department of Machine Learning, Moffitt Cancer Center & Research Institute, Tampa, FL, USA. aakash.tripathi@moffitt.org.
NPJ digital medicine
|October 23, 2025
概括
协调性ONcology生物医疗嵌入编码器 (HONeYBEE) 框架将各种癌症数据统一到患者嵌入中. 临床数据嵌入在分类和患者检索中实现了高准确度,增强了瘤学研究.
科学领域:
- 在瘤学瘤学.
- 生物医学信息学 生物医学信息学
- 人工智能在医学中的应用
背景情况:
- 整合多模式生物医学数据对于推进瘤学研究至关重要.
- 现有的框架往往难以统一不同的数据类型,如临床,成像和分子资料.
- 开发强大的患者级别表示是癌症预测建模的关键.
研究的目的:
- 引入协调的ONcology生物医学嵌入编码器 (HONeYBEE),这是一个开源框架,用于多式联络瘤数据集成.
- 为各种瘤学应用产生统一的患者级嵌入.
- 评估临床和多式联络嵌入在生存预测和患者相似性检索等任务中的表现.
主要方法:
- 利用特定领域的基础模型和融合策略来处理临床数据 (结构化/非结构化),整片图像,放射学扫描和分子资料.
- 创建了统一的患者级嵌入.
- 在癌症基因组图谱 (TCGA) 数据集 (>11,400名患者,33种癌症类型) 上评估了嵌入.
- 用于临床文本表示的通用和专用大型语言模型的性能比较.
主要成果:
- 临床嵌入显示出优异的单一模式性能:98.5%的癌症类型分类精度和96.4%的患者检索精度@10.
- 临床嵌入在大多数癌症类型中实现了最高的生存预测一致性指数.
- 多式融合提供了互补的好处,改善了超出特定癌症的临床特征单独的整体生存预测.
- 一般用途的大型语言模型 (例如,Qwen3) 在临床文本表示方面表现优于专业医疗模型,微调可以提高异质数据的性能.
结论:
- 蜜蜂有效地整合了瘤学的多式联络生物医学数据,产生了强大的患者级嵌入.
- 临床数据嵌入对于分类和检索任务非常有效,作为一个强大的基线.
- 多模式数据集成和先进的语言模型为提高预测准确性和揭示癌症研究中的新见解提供了巨大的潜力.
相关概念视频
Mouse Models of Cancer Study
6.4K
Mice have long served as models for studying human biology and pathology because of their phylogenetic and physiological similarity with humans. They are also easy to maintain and breed in the laboratory, and hence, many inbred strains are now available for research. Studies on mice have contributed immeasurably to our understanding of cancer biology.
The development of transgenic, knockout, and knock-in mice has led to an exponential increase in their use as model organisms in research,...
The development of transgenic, knockout, and knock-in mice has led to an exponential increase in their use as model organisms in research,...
6.4K
Tumor Immunotherapy
1.7K
Immunotherapy is a treatment that boosts or manipulates the immune system to fight diseases, including cancer. For instance, by stimulating an immune response through vaccinations against viruses that cause cancers, like hepatitis B virus and human papillomavirus, these diseases can be prevented. Nonetheless, some cancer cells can avoid the immune system due to their rapid mutation and division. The immune response to many cancers involves three phases: elimination, equilibrium, and escape.
1.7K
Adaptive Mechanisms in Cancer Cells
6.9K
Cancer cells accumulate genetic changes at an abnormally rapid rate due to the defects in the DNA repair mechanisms. From an evolutionary perspective, such genetic instability is advantageous for cancer development. Mutant cell lines accumulate a series of beneficial mutations that contribute to their progression into cancer.
Some of the advantages that cancer cells have on normal cells include - enhanced ability to divide without terminally differentiating, induce new blood vessel formation,...
Some of the advantages that cancer cells have on normal cells include - enhanced ability to divide without terminally differentiating, induce new blood vessel formation,...
6.9K

