A New Multiple Imputation Method for High-Dimensional Neuroimaging Data

IF 3.5 2区 医学 Q1 NEUROIMAGING
Tong Lu, Peter Kochunov, Chixiang Chen, Hsin-Hsiung Huang, L. Elliot Hong, Shuo Chen
{"title":"A New Multiple Imputation Method for High-Dimensional Neuroimaging Data","authors":"Tong Lu,&nbsp;Peter Kochunov,&nbsp;Chixiang Chen,&nbsp;Hsin-Hsiung Huang,&nbsp;L. Elliot Hong,&nbsp;Shuo Chen","doi":"10.1002/hbm.70161","DOIUrl":null,"url":null,"abstract":"<p>Missing data are a prevalent challenge in neuroimaging, with significant implications for downstream statistical analysis. Neglecting this issue can introduce bias and lead to erroneous inferential conclusions, making it crucial to employ appropriate statistical methods for handling missing data. Although the multiple imputation is a widely used technique, its application in neuroimaging is severely hindered by the high dimensionality of neuroimaging data, and the substantial computational demands. To tackle the critical computational challenges, we propose a novel approach, <b>H</b>igh d<b>i</b>mensional <b>M</b>ultiple Imput<b>a</b>tion (HIMA), based on Bayesian models specifically designed for large-scale neuroimaging datasets. HIMA introduces a new computational strategy to sample large covariance matrices based on a robustly estimated posterior mode, significantly improving both computational efficiency and numerical stability. To assess the effectiveness of HIMA, we conducted extensive simulation studies and real-data analysis from a Schizophrenia brain imaging dataset with around 1000 voxels. HIMA showcases a remarkable reduction of computational burden, for example, 1 hour by HIMA versus 800 hours by classic multiple imputation packages. HIMA also demonstrates improved precision and stability of imputed data.</p>","PeriodicalId":13019,"journal":{"name":"Human Brain Mapping","volume":"46 5","pages":""},"PeriodicalIF":3.5000,"publicationDate":"2025-03-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11926575/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Human Brain Mapping","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/hbm.70161","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"NEUROIMAGING","Score":null,"Total":0}
引用次数: 0

Abstract

Missing data are a prevalent challenge in neuroimaging, with significant implications for downstream statistical analysis. Neglecting this issue can introduce bias and lead to erroneous inferential conclusions, making it crucial to employ appropriate statistical methods for handling missing data. Although the multiple imputation is a widely used technique, its application in neuroimaging is severely hindered by the high dimensionality of neuroimaging data, and the substantial computational demands. To tackle the critical computational challenges, we propose a novel approach, High dimensional Multiple Imputation (HIMA), based on Bayesian models specifically designed for large-scale neuroimaging datasets. HIMA introduces a new computational strategy to sample large covariance matrices based on a robustly estimated posterior mode, significantly improving both computational efficiency and numerical stability. To assess the effectiveness of HIMA, we conducted extensive simulation studies and real-data analysis from a Schizophrenia brain imaging dataset with around 1000 voxels. HIMA showcases a remarkable reduction of computational burden, for example, 1 hour by HIMA versus 800 hours by classic multiple imputation packages. HIMA also demonstrates improved precision and stability of imputed data.

Abstract Image

一种新的高维神经影像数据多重输入方法。
数据缺失是神经影像学中普遍存在的挑战,对下游统计分析具有重要意义。忽略这个问题可能会引入偏见并导致错误的推断结论,因此使用适当的统计方法来处理丢失的数据至关重要。虽然多重插值是一种广泛应用的技术,但由于神经影像学数据的高维数和大量的计算需求,严重阻碍了其在神经影像学中的应用。为了解决关键的计算挑战,我们提出了一种新的方法,高维多重插值(HIMA),基于专门为大规模神经成像数据集设计的贝叶斯模型。HIMA引入了一种新的基于鲁棒估计后验模式的大协方差矩阵采样计算策略,显著提高了计算效率和数值稳定性。为了评估HIMA的有效性,我们进行了广泛的模拟研究和来自约1000体素的精神分裂症脑成像数据集的实际数据分析。HIMA显著地减少了计算负担,例如,HIMA减少了1小时,而经典的多重输入包减少了800小时。HIMA还证明了输入数据的精度和稳定性的提高。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Human Brain Mapping
Human Brain Mapping 医学-核医学
CiteScore
8.30
自引率
6.20%
发文量
401
审稿时长
3-6 weeks
期刊介绍: Human Brain Mapping publishes peer-reviewed basic, clinical, technical, and theoretical research in the interdisciplinary and rapidly expanding field of human brain mapping. The journal features research derived from non-invasive brain imaging modalities used to explore the spatial and temporal organization of the neural systems supporting human behavior. Imaging modalities of interest include positron emission tomography, event-related potentials, electro-and magnetoencephalography, magnetic resonance imaging, and single-photon emission tomography. Brain mapping research in both normal and clinical populations is encouraged. Article formats include Research Articles, Review Articles, Clinical Case Studies, and Technique, as well as Technological Developments, Theoretical Articles, and Synthetic Reviews. Technical advances, such as novel brain imaging methods, analyses for detecting or localizing neural activity, synergistic uses of multiple imaging modalities, and strategies for the design of behavioral paradigms and neural-systems modeling are of particular interest. The journal endorses the propagation of methodological standards and encourages database development in the field of human brain mapping.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信