Fourteen patients with stroke (age: 56.0 ± 13.7, M: F=9:5) were recruited. Table 1 shows the demographic information of patients with stroke. Twenty volunteers (6 males, 14 females) aged 20-28 years (mean age: 22.4 ± 2.41 years) participated in the experiment.
Table 1. Demographic Information of Patients With Stroke
ID
|
Age
|
Gender
|
Stroke type
|
Stroke location
|
Hemiparetic side
|
Time since stroke onset (months)
|
Brunnstrom-stage (arm)
|
Brunnstrom-stage (hand)
|
1
|
25
|
M
|
Hemorrhagic
|
Basal ganglia
|
Left
|
12
|
3
|
3
|
2
|
71
|
M
|
Ischemic
|
Corona radiata
|
Right
|
4
|
3
|
3
|
3
|
49
|
M
|
Hemorrhagic
|
Basal ganglia, corona radiata
|
Right
|
12
|
3
|
3
|
4
|
46
|
M
|
Hemorrhagic
|
Basal ganglia
|
Right
|
4
|
2
|
2
|
5
|
52
|
F
|
Hemorrhagic
|
Fronto-temporal
|
Left
|
13
|
4
|
3
|
6
|
75
|
F
|
Ischemic
|
Basal ganglia, corona radiata
|
Left
|
14
|
2
|
2
|
7
|
70
|
F
|
Ischemic
|
Middle cerebral artery territory
|
Left
|
9
|
2
|
2
|
8
|
69
|
M
|
Ischemic
|
Basal ganglia, corona radiata
|
Left
|
9
|
3
|
3
|
9
|
42
|
M
|
Hemorrhagic
|
Basal ganglia
|
Left
|
8
|
3
|
3
|
10
|
49
|
M
|
Hemorrhagic
|
Fronto-parietal
|
Right
|
7
|
3
|
3
|
11
|
55
|
F
|
Hemorrhagic
|
Basal ganglia
|
Left
|
7
|
3
|
3
|
12
|
54
|
M
|
Hemorrhagic
|
Basal ganglia
|
Left
|
16
|
3
|
3
|
13
|
61
|
F
|
Hemorrhagic
|
Corona radiata
|
Left
|
5
|
4
|
3
|
14
|
66
|
M
|
Hemorrhagic
|
Basal ganglia, thalamus
|
Right
|
13
|
4
|
3
|
Mu suppression in the M1 and parietal cortex
The mixed-effects ANOVA on mu suppression (Table 2) showed a significant main effect between the two conditions (MVF vs no-MVF) on M1 (p<0.001) and the parietal cortex (p=0.003), and there was a significant interaction effect among the three factors (MVF condition * Group * Hemisphere) on the parietal cortex (p=0.006). Other comparisons did not reveal any significant effects.
The mu suppression values for MVF and no-MVF in the ipsilesional M1 were -0.953±1.699 and 0.166±1.470 for healthy controls and -1.555±0.870 and -0.317±0.730 for patients with stroke, respectively (Figure 2a). Both groups showed stronger mu suppression in the MVF than in the no-MVF in the ipsilesional M1 (p=0.003 and p=0.006 for healthy controls and patients with stroke, respectively). Meanwhile, those for the contralesional M1 region were -0.795±2.527 and 0.025±1.671 for healthy controls, and -1.246±1.230 and -0.419±0.924 for patients with stroke in the MVF and no-MVF groups, respectively, in which only healthy controls showed significantly stronger suppression in the MVF group than in the no-MVF group (p=0.041).
The mu suppression values in the ipsilesional parietal cortices were -0.162±0.428 and -1.574±1.902 for healthy controls and patients with stroke, respectively, in the MVF group, and 0.125±0.773 and 0.050±1.113 for healthy controls and patients with stroke, respectively, in the no-MVF group (Figure 2b). In addition, the mu suppression values in the contralesional parietal cortex were -0.370±2.809 for healthy controls and -0.971±2.140 for patients with stroke in the MVF condition, and 0.776±1.944 for the healthy control and -0.254±1.496 for patients with stroke in the no-MVF condition. In patients with stroke, mu suppression in the ipsilesional parietal cortex showed stronger suppression in the MVF (p=0.009). In contrast, in healthy controls, mu suppression in the contralesional parietal cortex was more suppressed in the MVF (p=0.036).
Table 2. Results of Mixed Effects ANOVA on Mu Suppression.
Factors
|
M1
|
Parietal cortex
|
F
|
P
|
F
|
P
|
MVF condition (MVF vs no-MVF)
|
25.489
|
<.001
|
10.522
|
.003
|
Group (Stroke vs healthy controls)
|
1.205
|
.281
|
3.742
|
.062
|
Hemispherea
|
0.078
|
.782
|
0.350
|
.558
|
MVF condition*Group
|
0.025
|
.875
|
0.609
|
.441
|
MVF condition*Hemisphere
|
1.701
|
.201
|
0.006
|
.939
|
Hemisphere*Group
|
0.056
|
.815
|
0.013
|
.910
|
MVF condition*Group*Hemisphere
|
0.042
|
.839
|
8.513
|
.006
|
Abbreviations: MVF, mirror visual feedback.
aIpsilesional (ipsilateral to the moving hand in controls) vs contralesional (contralateral to the moving hand in controls).
Mu coherence between M1 and parietal cortex (intrahemispheric, interhemispheric)
For mu coherence between the ipsilateral M1 and parietal cortex (IntraMPC), the mixed effect ANOVA (Table 3) showed significant main effects in the MVF condition (p=0.001), group (p<0.001), and hemisphere (p<0.001), and there were significant interaction effects in the MVF condition * group (p=0.016) and Hemisphere * Group (p<0.001).
Intra-MPC in patients with stroke was stronger in no-MVF than in MVF in both ipsilesional (p=0.013) and contralesional hemispheres (p=0.041), while no difference was observed in healthy controls (Figure 3).
IntraMPC in the ipsilesional hemisphere was stronger in patients with stroke than in healthy controls, regardless of the MVF condition (p<0.001). At the same time, there was no significant difference in the contralesional hemisphere between the stroke and control groups. In addition, IntraMPC was stronger in the contralesional hemisphere (contralateral to the moving hand) than in the ipsilesional hemisphere, regardless of the MVF condition. At the same time, patients with stroke did not show any differences according to the hemisphere side.
Interhemispheric mu coherence between the M1 cortices (InterMC) was significantly stronger in healthy controls than in patients with stroke (p=0.032). At the same time, the mixed effect ANOVA model revealed no significant effect of MVF on IntraMC and IntraPC in patients with stroke and healthy controls.
Table 3. Results of Mixed Effects ANOVA on Mu Coherence.
Factors
|
Mu coherence between ipsilateral
M1-parietal cortex
|
|
F
|
P
|
MVF condition (MVF vs no-MVF)
|
12.171
|
.001
|
Group (Stroke vs healthy controls)
|
42.996
|
<.001
|
Hemispherea
|
47.487
|
<.001
|
MVF condition*Group
|
6.501
|
.016
|
MVF condition*Hemisphere
|
0.444
|
.510
|
Hemisphere*Group
|
19.196
|
<.001
|
MVF condition*Group*Hemisphere
|
1.281
|
.266
|