在精神分裂症诊断中的多站点域概括的原型表示学习
IEEE transactions on bio-medical engineering
|January 28, 2026
概括
这项研究引入了一种新的域泛化框架,用于使用大脑功能网络诊断精神分裂症,显著提高了不同站点的分类准确性. 该方法通过学习站点不变特征并利用原型学习来提高模型概括性,以获得强大的诊断性能.
科学领域:
- 神经科学是一个神经科学.
- 机器学习 机器学习
- 医疗成像医学成像
背景情况:
- 来自fMRI的大脑功能网络 (BFNs) 对于诊断精神疾病如精神分裂症 (SZ) 至关重要.
- 多站点fMRI数据中的站点诱导的分布转移阻碍了分类模型的泛化.
- 现有的域泛化 (DG) 方法往往会失败,因为它假定跨站点存在一致的类结构,忽视了类内部的多样性.
研究的目的:
- 开发一个强大的域泛化 (DG) 框架,使用多站点BFNs对精神分裂症进行分类.
- 为了克服fMRI数据中场所诱导的分布转移的挑战.
- 改进诊断模型对未见域的概括性能.
主要方法:
- 使用变压器编码器从BFNs中提取歧视性主体级表示.
- 一个具有希尔伯特-施密特独立标准 (HSIC) 正规化的站点独立模块强制执行站点不变特征学习.
- 原型学习采用Sinkhorn匹配,指数移动平均 (EMA) 更新和最大概率估计 (MLE) 损失精细的功能对原型匹配.
主要成果:
- 拟议的GD框架在两个独立的SZ数据集 (88.89%±2.22%和86.05%±1.64%) 上实现了优异的分类性能,超过了7个GD,6个域调整 (DA),6个多站点和6个最先进的方法.
- 废除研究证实了MLE,对比和对齐损失对性能提升的重大贡献.
- 该方法确定了歧视性时区域,提供了对精神分裂症背后的神经机制的见解.
结论:
- 开发的基于变压器的GD框架有效地解决了针对精神分裂症诊断的多站点fMRI数据的站点诱导的分布转移.
- 原型学习和站点不变性约束的整合导致了增强的概括性和强大的分类性能.
- 这些发现为精神分裂症的神经生物学基础提供了宝贵的见解,并突出了高级机器学习技术在精神病学研究中的潜力.
相关概念视频
Schizophrenia
934
Schizophrenia, a term introduced by Swiss psychiatrist Eugen Bleuler in 1911, describes a severe psychological disorder marked by profound disruptions in attention, thought processes, language, emotion, and interpersonal relationships. The core feature of schizophrenia is psychosis — a state characterized by a fundamental detachment from reality. This disconnection manifests through distorted logic, impaired perception, and atypical behavior, severely affecting the lives of those...
934
Concepts and Prototypes
538
The human nervous system handles vast amounts of information by translating sensory stimuli into neural impulses, which the brain processes, creating thoughts expressed through language or stored as memories. The brain also synthesizes information from emotions and memories, which significantly influence thoughts and behaviors. This intricate process creates a comprehensive mental picture.
The brain organizes this information using concepts, which are mental categories grouping linguistic data,...
The brain organizes this information using concepts, which are mental categories grouping linguistic data,...
538
Biological Causes of Schizophrenia
596
Schizophrenia, a severe psychiatric disorder, arises from a complex interplay of biological factors, including genetic predisposition, structural brain abnormalities, neurotransmitter dysregulation, and developmental irregularities. These factors collectively contribute to the onset and progression of the disorder, which typically manifests in late adolescence or early adulthood.
Genetic Factors in Schizophrenia
The genetic basis of schizophrenia is strongly supported by family and twin...
Genetic Factors in Schizophrenia
The genetic basis of schizophrenia is strongly supported by family and twin...
596
Generalized Hooke's Law
2.7K
The generalized Hooke's Law is a broadened version of Hooke's Law, which extends to all types of stress and in every direction. Consider an isotropic material shaped into a cube subjected to multiaxial loading. In this scenario, normal stresses are exerted along the three coordinate axes. As a result of these stresses, the cubic shape deforms into a rectangular parallelepiped. Despite this deformation, the new shape maintains equal sides, and there is a normal strain in the direction of the...
2.7K
Generalized Anxiety Disorder
692
Generalized Anxiety Disorder (GAD) is a chronic condition characterized by excessive and uncontrollable worry that persists for at least six months, significantly interfering with daily functioning. Unlike situational anxiety, which arises in response to specific stressors, GAD often occurs without a clear cause. Individuals may experience disproportionate worry about work, health, or relationships. For instance, a person might continuously fear poor health despite normal medical evaluations or...
692
State Space Representation
566
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...
566


