Tensorial extensions of independent component analysis for multisubject FMRI analysis
Tensorial extensions of independent component analysis for multisubject FMRI analysis
复制标题
DOI：
10.1016/j.neuroimage.2004.10.043
复制
发表时间：
2005-03-01
期刊：
影响因子：
5.7
通讯作者：
Smith, SM
中科院分区：
文献类型：
--
作者：
Beckmann, CF;Smith, SM
我们通过扩展单会话概率独立成分分析模型（PICA; Beckmann and Smith，2004）讨论了多主题或多会话FMRI数据的无模型分析。IEEE Trans. on Medical Imaging，23（2）137-152）到更高维度。这导致了三向分解，其表示数据中存在的不同信号和伪影的时间、空间和受试者相关变化。该技术来自平行因子分析（PARAFAC; Harshman和Lundy，1984），并与之进行了比较。多模数据分析的研究方法，第5章，第122-215页。Praeger，纽约）。使用模拟数据，以及多会话和多学科功能磁共振成像研究的数据，我们证明了张量PICA方法是能够有效和准确地提取感兴趣的信号在空间，时间和主题/会话域。最终的分解在更高的准确性、减少不同估计源之间的干扰（减少串扰）、鲁棒性（针对数据与建模假设的偏差和针对过拟合）和计算速度方面改进了PARAFAC结果。在真实的FMRI“激活”数据上，张量PICA方法能够提取合理的激活图、时间过程和会话/受试者模式，以及提供对感兴趣的附加过程（例如图像伪影或次级激活模式）的丰富描述。由此产生的数据分解提供了简单而有用的表示多学科/多会话功能磁共振成像数据，可以帮助解释和优化组功能磁共振成像研究超出了什么可以实现基于模型的分析技术。(C)2004年爱思唯尔公司All rights reserved.
We discuss model-free analysis of multisubject or multisession FMRI data by extending the single-session probabilistic independent component analysis model (PICA; Beckmann and Smith, 2004. IEEE Trans. on Medical Imaging, 23 (2) 137-152) to higher dimensions. This results in a three-way decomposition that represents the different signals and artefacts present in the data in terms of their temporal, spatial, and subject-dependent variations. The technique is derived from and compared with parallel factor analysis (PARAFAC; Harshman and Lundy, 1984. In Research methods for multimode data analysis, chapter 5, pages 122-215. Praeger, New York). Using simulated data as well as data from multisession and multisubject FMRI studies we demonstrate that the tensor PICA approach is able to efficiently and accurately extract signals of interest in the spatial, temporal, and subject/session domain. The final decompositions improve upon PARAFAC results in terms of greater accuracy, reduced interference between the different estimated sources (reduced cross-talk), robustness (against deviations of the data from modeling assumptions and against overritting), and computational speed. On real FMRI 'activation' data, the tensor PICA approach is able to extract plausible activation maps, time courses, and session/subject modes as well as provide a rich description of additional processes of interest such as image artefacts or secondary activation patterns. The resulting data decomposition gives simple and useful representations of multisubject/multisession FMRI data that can aid the interpretation and optimization of group FMRI studies beyond what can be achieved using model-based analysis techniques. (C) 2004 Elsevier Inc. All rights reserved.