整合中央切片解剖学和MRF映射的放射力学动态,用于肺腺癌亚型化
Weiwei Shi1, He Ren2, Chengcheng Fan2
1Shanghai YangZhi Rehabilitation Hospital (Shanghai Sunshine Rehabilitation Center), School of Medicine, Tongji University, Shanghai, China.
Quantitative imaging in medicine and surgery
|February 11, 2026
概括
这项研究引入了一种新的动态放射性融合网络 (DRFN),用于肺腺癌亚型. DRFN将CT成像与放射学数据集成在一起,实现更高的准确性和可解释性,以改善临床诊断.
科学领域:
- 医学成像分析分析 医学成像分析
- 人工智能在瘤学中的应用
- 无线电学和深度学习
背景情况:
- 肺腺癌是一种主要的肺癌亚型,需要精确的分类来进行有效的临床管理.
- 目前基于计算机断层扫描 (CT) 的方法难以在相邻切片上捕捉正在演变的病变特征.
- 需要先进的方法来整合静态解剖细节与动态放射性变化.
研究的目的:
- 开发和验证一个新的深度学习框架,动态放射学融合网络 (DRFN),用于肺腺癌亚型.
- 通过将中央切片解剖与动态放射学融合,提高基于CT的肺腺癌分类的准确性和可解释性.
- 将DRFN的性能与肺腺癌亚型化现有的深度学习模型进行比较.
主要方法:
- 一个多中心的回顾性研究,利用来自培训,内部测试和外部验证队伍的CT病变.
- 提取中央切片图像 (IMG1) 和多层放射性序列,然后使用LASSO进行特征选择.
- 开发DRFN,一个双通道网络,将IMG1与BiLSTM-Attention衍生的马尔科夫随机场 (MRF) 地图融合在一起,并与CNN和BiLSTM-Attention基线进行比较.
主要成果:
- 在一个独立的测试队列中,DRFN实现了高的每类AUC:0.97 (最小侵入性腺癌),0.99 (腺癌 in situ) 和0.95 (侵入性腺癌).
- 类激活映射 (CAM) 显示了DRFN的重点是临床相关的特征,如病变核心,状边缘和血管结构.
- 在灵敏度,特异性和通用性方面,DRFN显著超过单通道CNN (AUCs:0.77-0.88) 和仅用于放射学的BiLSTM-Attention模型 (AUCs:0.88-0.94).
结论:
- DRFN框架有效地将静态解剖CT信息与动态放射学演变融合在一起,从而产生高分类准确度.
- 通过注意力机制,DRFN提供了透明的解释性,反映了放射科医生的诊断推理.
- 这种智能诊断辅助器在改善临床实践中的肺腺癌亚型方面显著有前途.
相关概念视频
Gross Anatomy of the Lungs
4.8K
The lungs are a pair of vital organs connected to the trachea via the left and right bronchi. The base of these organs meets the dome-shaped muscle known as the diaphragm. Encased by the pleurae, the lungs contact the mediastinum. The right lung is shorter yet wider, and has a larger volume than the left lung. The left lung has an indentation known as the cardiac notch. The superior region of the lungs is referred to as the apex, whereas the base is the lower region near the diaphragm. The...
4.8K
Anatomy of the Circulatory System
98.4K
The human circulatory system consists of blood, blood vessels that carry blood away from the heart, around the body, and back to the heart, and the heart itself, which acts as a central pump. The systemic circuit supplies blood to the whole body, the coronary circuit supplies blood to the heart, and the pulmonary circuit supplies blood flow between the heart and lungs.
98.4K
The Central Dogma
140.0K
Overview
140.0K
Measures of Central Tendency
21.5K
The "center" of a data set is also a way of describing location. The two most widely used measures of the "center" of the data are the mean (average) and the median. The words "mean" and "average" are often used interchangeably. The substitution of one word for the other is common practice. The technical term is "arithmetic mean" and "average" is technically a center location. However, in practice among non-statisticians,...
21.5K
Lung Capacity
56.4K
The air in the lungs is measured in volumes and capacities. Lung volume measures reflect the amount of air taken in, released, or left over after a lung function, like a single inhalation. Lung capacity measures are sums of two or more lung volume measures.
56.4K
Anatomy of the Heart
120.6K
The human heart is made up of three layers of tissue that are surrounded by the pericardium, a membrane that protects and confines the heart. The outermost layer, closest to the pericardium, is the epicardium. The pericardial cavity separates the pericardium from the epicardium. Beneath the epicardium is the myocardium, the middle layer, and the endocardium, the innermost layer. There are four chambers of the heart: the right atrium, the right ventricle, the left atrium, and the left ventricle.
120.6K


