一个多实例的学习网络与原型实例对抗对抗对宫病理学分级
Mingrui Ma1, Furong Luo2, Binlin Ma3
1School of Computer Science and Technology, Xinjiang University, Urumqi 830046, China; Affiliated Tumor Hospital of Xinjiang Medical University, Xinjiang 830011, China.
Medical image analysis
|December 18, 2025
概括
这项研究介绍了PacMIL,这是一个新的深度学习网络,用于部状细胞癌分级. 通过解决现有的多实例学习方法的模糊性,PacMIL提高了病理分级的准确性.
科学领域:
- 在瘤学瘤学.
- 计算病理学计算病理学
- 人工智能的人工智能
背景情况:
- 宫平细胞癌 (CSCC) 的病理分级对于瘤诊断至关重要.
- 当前的多实例学习 (MIL) 方法与来自差异化图像区域的模两可的分级模式作斗争.
- 这种模糊性限制了CSCC病理分级模型的准确性.
研究的目的:
- 开发一个先进的深度学习模型,用于准确的CSCC病理分级.
- 克服现有的MIL方法在处理模两可的分级模式方面的局限性.
- 在病理学图像中增强单个和多个差异化实例的表示学习.
主要方法:
- 提出了一个名为PacMIL的端到端多实例学习网络.
- 引入了一个非平衡学习算法来解决MIL表示不匹配的问题.
- 设计了一种具有注意力机制的原型实例对抗性对比 (PAC) 方法.
- 将对抗式对比学习和度量距离纳入优化目标.
主要成果:
- 帕克米尔获得了93.09%的精度 (mAcc) 和0.9802的AUC.
- 该模型在CSCC病理分类中显著超过了最先进的 (SOTA) 方法.
- 与现有方法相比,PacMIL表现出优越的代表性能力.
结论:
- 拟议的PacMIL模型为CSCC病理分级提供了增强的实用性.
- PacMIL有效地解决了病理图像分级中的模两可,提高了诊断准确度.
- 这项研究为计算病理学和癌症诊断提供了宝贵的工具.
相关概念视频
01:27Associative Learning
1.2K
Associative learning is a fundamental concept in behavioral psychology, wherein a connection is established between two stimuli or events, leading to a learned response. This process is critical in understanding how behaviors are acquired and modified. Conditioning, the mechanism through which associations are formed, can be divided into two main types: classical conditioning and operant conditioning, each elucidating different aspects of associative learning.
Classical conditioning, also known...
Classical conditioning, also known...
1.2K
01:22Classification of Epithelial Tissues: Overview
19.5K
Epithelial tissues are classified according to the shape of the cells and the number of cell layers formed. Cell shapes can be squamous (flattened and thin), cuboidal (square-like, as wide as it is tall), or columnar (rectangular, taller than it is wide). Additionally, the nucleus shape helps identify the type of epithelial cells. Squamous cells have flattened disc-shaped nuclei, cuboidal cells have spherical nuclei, and columnar cells have elongated nuclei.
Based on the number of cell layers,...
Based on the number of cell layers,...
19.5K