Heart Rate and Respiration Affect the Functional Connectivity of Default Mode Network in Resting-State Functional Magnetic Resonance Imaging
Heart Rate and Respiration Affect the Functional Connectivity of Default Mode Network in Resting-State Functional Magnetic Resonance Imaging
复制标题
DOI：
10.3389/fnins.2020.00631
复制
发表时间：
2020-06
影响因子：
4.3
通讯作者：
Akira Yoshikawa;Y. Masaoka;Masaki Yoshida;N. Koiwa;M. Honma;Keiko Watanabe;Satomi Kubota;Iizuka Natsuko;M. Ida;M. Izumizaki
中科院分区：
文献类型：
--
作者：
Akira Yoshikawa;Y. Masaoka;Masaki Yoshida;N. Koiwa;M. Honma;Keiko Watanabe;Satomi Kubota;Iizuka Natsuko;M. Ida;M. Izumizaki
越来越多的脑成像研究表明，在人类的情感和认知任务中，脑区之间存在功能连接（FC）。然而，情绪伴随着心率和呼吸等生理参数的变化。这些变化可能会影响依赖于血氧水平的信号，以及大脑区域之间的连接。本研究旨在利用静息状态功能磁共振成像（rs-fMRI）阐明生理噪声对默认模式网络相关区域之间连通性的影响。健康成年志愿者（年龄19-51岁，平均26.9±9.1岁，男8名，女8名）使用临床3T扫描仪（Siemens MAGNETOM Trio a Tim System）进行10分钟的rs-fMRI，同时记录呼吸和心输出量。随后使用DRIFTER工具箱从获得的fMRI数据中去除生理噪声信号。使用DPARSF对与默认模式网络相关的区域之间的FC进行图像处理和分析。采用网络统计（Network-Based Statistic， NBS）方法对DMN及DMN相关区域的功能连接体进行三组比较：无生理性噪声校正、有心脏噪声校正、有心脏和呼吸噪声校正。NBS分析确定了36个网络在FC矩阵的三种情况下存在显著差异。事后比较显示三种情况之间没有差异，这表明三种情况都有相同的网络。在36个网络中，8个网络的FC强度在生理噪声校正下得到了修正。左右额叶前部内侧区域之间的连通性增加了连通性的强度。这些区域位于大脑内侧半球，靠近大脑半球的矢状窦和动脉，提示内侧额叶区可能对靠近动脉的心律敏感。其他网络观察到颞区，并通过去除生理噪声显示其连通性强度下降，这表明生理噪声，特别是呼吸，可能在静息状态时对颞区BOLD信号敏感。颞叶与焦虑相关的呼吸变化（Masaoka and Homma, 2000）、言语处理和呼吸感觉高度相关。这些因素可能影响rs-fMRI信号敏感性。
A growing number of brain imaging studies show functional connectivity (FC) between regions during emotional and cognitive tasks in humans. However, emotions are accompanied by changes in physiological parameters such as heart rate and respiration. These changes may affect blood oxygen level-dependent signals, as well as connectivity between brain areas. This study aimed to clarify the effects of physiological noise on the connectivity between areas related to the default mode network using resting-state functional magnetic resonance imaging (rs-fMRI). Healthy adult volunteers (age range: 19–51 years, mean age: 26.9 ± 9.1 years, 8 males and 8 females) underwent rs-fMRI for 10 min using a clinical 3T scanner (MAGNETOM Trio A Tim System, Siemens) with simultaneously recorded respiration and cardiac output. Physiological noise signals were subsequently removed from the acquired fMRI data using the DRIFTER toolbox. Image processing and analysis of the FC between areas related to the default mode network were performed using DPARSF. Network-Based Statistic (NBS) analysis of the functional connectome of the DMN and DMN-related area was used to perform three groups of comparison: without physiological noise correction, with cardiac noise correction, and with cardiac and respiratory noise correction. NBS analysis identified 36 networks with significant differences in three conditions in FC matrices. Post hoc comparison showed no differences between the three conditions, indicating that all three had the same networks. Among the 36 networks, strength of FC of 8 networks was modified under physiological noise correction. Connectivity between left and right anterior medial frontal regions increased strength of connectivity. These areas are located on the medial cerebral hemisphere, close to the sagittal sinus and arteries in the cerebral hemispheres, suggesting that medial frontal areas may be sensitive to cardiac rhythm close to arteries. The other networks observed temporal regions and showed a decrease in their connectivity strength by removing physiological noise, indicating that physiological noise, especially respiration, may be sensitive to BOLD signal in the temporal regions during resting state. Temporal lobe was highly correlated with anxiety-related respiration changes (Masaoka and Homma, 2000), speech processing, and respiratory sensation. These factors may affect the rs-fMRI signaling sensitivity.