| 제목 |
Different intrinsic functional coupling and effect of dual-task interference in Parkinson’s disease |
| 소속 |
Seoul National University Hospital, Department of Rehabilitation Medicine1, Seoul National University Hospital, Biomedical Research Institute2, Seoul National University College of Medicine, Department of Rehabilitation Medicine3, Dong-A University College of Medicine, Department of Physical Medicine and Rehabilitation4, Chung-Ang University College of Medicine, Department of Physical Medicine and Rehabilitation5 |
| 저자 |
Eunkyung Kim1,2*, Heejae Kim1, Seo Jung Yun1, Min-Gu Kang1,4, Hyun Iee Shin1,5, Byung-Mo Oh1,3, Han Gil Seo1,3† |
| 사사 |
Grants
This study is supported by the Translational Research Program for Rehabilitation Robots (NRCTR-EX18009), National Rehabilitation Center, Ministry of Health and Welfare, Korea and supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (2020R1C1C1012785).
References
[1] C.F. Beckmann, C.E. Mackay, N. Filippini, S.M. Smith, Group comparison of resting-state FMRI data using multi-subject ICA and dual regression, Neuroimage 47(Suppl 1) (2009) S148.
[2] G. Dirnberger, M. Jahanshahi, Executive dysfunction in Parkinson's disease: a review, J. Neuropsychol. 7(2) (2013) 193-224.
[3] J. Rowe, K.E. Stephan, K. Friston, R. Frackowiak, A. Lees, R. Passingham, Attention to action in Parkinson’s disease: impaired effective connectivity among frontal cortical regions, Brain 125(2) (2002) 276-289.
|
Introduction
To examine changes of functional brain network coupling including the frontoparietal control network (FPN), dorsal attention network (DAN), motor network (MN), and lateral motor network (LMN) in individuals with Parkinson’s disease (PD) compared to controls associated with dual-task interference.
Methods
A total of 28 participants of idiopathic PD (Hoehn & Yahr stage of 2.5 or 3, Mini-Mental State Examination score ≥ 24, 68.9±8.8 years) and their age- and sex- matched healthy controls (67.2±8.5 years) were included. Resting state functional magnetic resonance imaging (rsfMRI) data was acquired using a 3-T scanner. The participants performed a 10-meter walking test or timed up and go test under dual-task conditions (cognitive and physical tasks). Dual-task interference was calculated as a measure of gait automaticity.
After preprocessing the rsfMRI data, intrinsic functional sub-networks were extracted by independent component analysis [1] using all individual data. Among the extracted group-specific 20 sub-networks, 5 sub-networks, including the left and right FPNs, DAN, MN and LMN were identified. Subject-specific time-course and spatial component from the sub-networks were estimated by applying dual regression approach.
Functional coupling among the 5 sub-networks were estimated by Pearson’s correlation analysis using the subject-specific time-course data and compared between the groups using permutation test. Correlation between the functional coupling and dual-task interference was also estimated in each group, respectively.
Results
Except the right FPN, interplay among the 4 functional sub-networks was different between the groups (false discovery rate correction (FDR) was applied using threshold-free cluster enhancement (TFCE) < .05, Fig 1). In particular, functional coupling between left FPN and MN/LMN was increased in PD, while functional coupling between DAN and MN/LMN was decreased in PD compared to controls. Cognitive dual-task interference was negatively correlated with the coupling between left FPN and LMN in controls. A significant positive correlation between left FPN and LMN was observed in PD when excluding an outlier (P <.05, Fig 2).
Discussion
The results suggested different functional coupling of left FPN and DAN over motor-related networks in PD, although executive control and goal-directed attention were all impaired in PD [2] [3]. Individuals with PD might have a weak association of goal-directed attention over motor function compared to controls. On the other hand, the enhanced executive control of motor function observed in PD might reflect brain change to overcome gait disturbance by strengthening cognitive control of movement. The finding of enhanced executive control of motor function related with improved gait automaticity in PD supported this explanation. The results might help unravel neural mechanism of gait function in PD and may be useful in future study for improving gait automaticity in PD. |
|
Fig. 1. Results of group comparison for inter-component correlation between dorsal attention network (DAN), left frontoparietal network (FPN-L), motor network (MN), and lateral motor network (LMN). Elements of upper diagonal matrix represent increased functional coupling in the patient group compared to the control group while those of lower diagonal matrix represent decreased functional coupling in the patient group compared to the control group. In each element in the upper and lower diagonal matrix, the plots represent z value of corresponding inter-component correlation in each individual and in each group. The difference between group is statistically significant by using a permutation test, and the original t-value is described right above the plot. The asterisk represents the case in which false discovery rate is not corrected (P < .05). The intrinsic functional sub-network is displayed on the diagonal elements of the matrix.
Fig. 2. Relationship between dual task interference and inter-component correlation of left frontoparietal network (FPN-L) and lateral motor network (LMN). (A) Visualization of the FPN-L and LMN. (B) plot of z value of inter-component correlation between FPN-L and LMN in each individual and each group. (C) plot of z value of inter-component correlation (x-axis) and dual task interference cognitive performance (y-axis) in controls (black) and patients (blue). Regression lines represented negative (black solid line for controls) and positive (green dotted line for patients) correlations.
|
