一个双分支网络用于医疗图像中的病变细分,使用状态空间模型
Hao Chen1,2, Byung-Won Min2, Haifei Zhang1
1School of Information Engineering, Nantong Institute of Technology, Nantong, China.
Quantitative imaging in medicine and surgery
|December 10, 2025
概括
这项研究引入了一种新的双分支网络,将状态空间模型和卷积网络结合起来,以改善医疗损伤细分. 新方法增强了特征提取,导致临床应用中更准确的细分结果.
科学领域:
- 医疗成像医学成像
- 计算机视觉 计算机视觉
- 人工智能的人工智能
背景情况:
- 在医学图像中精确的病变细分对于诊断和治疗至关重要.
- 目前的方法难以捕捉本地和全球图像特征,限制了细分的准确性.
- 需要先进的模型,可以有效地整合各种特征信息,以精确识别病变.
研究的目的:
- 开发一个新的双分支网络,以加强医疗损伤细分.
- 改进本地和全球特征的提取,以获得更高的细分精度.
- 提供一个计算效率高,适应性强的模型,用于对各种大小的病变进行细分.
主要方法:
- 一个双分支编码器,集成一个视觉状态空间编码器 (用于远程依赖) 和一个残余网络 (用于本地特征).
- 一个轻量级的多尺度深度可分离的卷积块,用于精致的特征表示.
- 功能融合,然后用解码器进行高精度损伤细分.
主要成果:
- 拟议的模型在Kaggle_3M和Kvasir-SEG数据集上比最先进的方法取得了更高的性能.
- 在Kaggle_3M上获得0.9140的子相似系数 (Dice) 和0.0800的虚假负率 (FNR).
- 在Kvasir-SEG上获得了0.9173的子和0.0788的FNR,证明了强大的跨数据集泛化.
结论:
- 这种新型的双分支网络有效地集成了状态空间模型和深度卷积网络,以改善损伤细分.
- 该模型能够捕获本地和全球特征,这为医学图像分析中的临床应用提供了显著的潜力.
- 这项工作为推进医疗图像细分技术提供了宝贵的见解.
相关概念视频
State Space Representation
499
The frequency-domain technique, commonly used in analyzing and designing feedback control systems, is effective for linear, time-invariant systems. However, it falls short when dealing with nonlinear, time-varying, and multiple-input multiple-output systems. The time-domain or state-space approach addresses these limitations by utilizing state variables to construct simultaneous, first-order differential equations, known as state equations, for an nth-order system.
Consider an RLC circuit, a...
Consider an RLC circuit, a...
499
State Space to Transfer Function
536
The conversion of state-space representation to a transfer function is a fundamental process in system analysis. It provides a method for transitioning from a time-domain description to a frequency-domain representation, which is crucial for simplifying the analysis and design of control systems.
The transformation process begins with the state-space representation, characterized by the state equation and the output equation. These equations are typically represented as:
The transformation process begins with the state-space representation, characterized by the state equation and the output equation. These equations are typically represented as:
536


