Table 2.
Comparison of Network Metrics Computed Using Different Microstructural Weightings Between Multiple Sclerosis Patients and Healthy Controls
| Network metrics | Healthy controls | MS patients | Adjusted p value metrics | Adjusted p value model | |
|---|---|---|---|---|---|
| FA | Efficiency | 0.410 ± 0.014 | 0.394 ± 0.021 | 0.639 | 1.000 |
| Modularity | 0.065 ± 0.010 | 0.078 ± 0.020 | 0.477 | 0.812 | |
| Clustering coefficient | 0.382 ± 0.015 | 0.367 ± 0.020 | 0.639 | 1.000 | |
| Mean strength | 27.300 ± 1.400 | 25.500 ± 2.380 | 0.639 | 1.000 | |
| -ln(MD) | Efficiency | 6.030 ± 0.087 | 5.900 ± 0.179 | 0.020 | 0.075 |
| Modularity | 0.080 ± 0.015 | 0.097 ± 0.027 | 0.606 | 1.000 | |
| Clustering coefficient | 5.810 ± 0.100 | 5.69 ± 0.160 | 0.218 | 0.812 | |
| Mean strength | 412.000 ± 13.900 | 393.000 ± 28.100 | 0.020 | 0.062 | |
| -ln(RD) | Efficiency | 6.320 ± 0.097 | 6.180 ± 0.196 | 0.049 | 0.124 |
| Modularity | 0.078 ± 0.015 | 0.095 ± 0.027 | 0.489 | 1.000 | |
| Clustering coefficient | 6.080 ± 0.110 | 5.950 ± 0.176 | 0.203 | 0.812 | |
| Mean strength | 432.000 ± 14.800 | 411.00 ± 30.000 | 0.049 | 0.130 | |
| ICVF | Efficiency | 0.499 ± 0.023 | 0.472 ± 0.034 | 0.017 | 0.057 |
| Modularity | 0.069 ± 0.014 | 0.088 ± 0.026 | 0.017 | 0.069 | |
| Clustering coefficient | 0.470 ± 0.023 | 0.444 ± 0.034 | 0.017 | 0.069 | |
| Mean strength | 33.500 ± 2.020 | 30.700 ± 3.370 | 0.017 | 0.050 | |
| -ln(ISOVF) | Efficiency | 2.120 ± 0.119 | 2.090 ± 0.131 | 1.000 | 1.000 |
| Modularity | 0.116 ± 0.014 | 0.131 ± 0.025 | 1.000 | 1.000 | |
| Clustering coefficient | 1.980 ± 0.113 | 1.950 ± 0.119 | 1.000 | 1.000 | |
| Mean strength | 141.000 ± 9.060 | 135.000 ± 12.500 | 1.000 | 1.000 | |
| INTRA | Efficiency | 0.486 ± 0.024 | 0.456 ± 0.038 | 0.028 | 0.077 |
| Modularity | 0.069 ± 0.014 | 0.087 ± 0.025 | 0.030 | 0.214 | |
| Clustering coefficient | 0.456 ± 0.024 | 0.426 ± 0.038 | 0.028 | 0.077 | |
| Mean strength | 32.500 ± 2.080 | 29.600 ± 3.600 | 0.028 | 0.077 | |
| -ln(EXTRAMD) | Efficiency | 5.660 ± 0.082 | 5.550 ± 0.161 | 0.174 | 0.261 |
| Modularity | 0.080 ± 0.015 | 0.098 ± 0.027 | 0.794 | 1.000 | |
| Clustering coefficient | 5.450 ± 0.094 | 5.360 ± 0.143 | 0.794 | 1.000 | |
| Mean strength | 387.000 ± 13.000 | 369.000 ± 26.000 | 0.174 | 0.261 | |
| -ln(EXTRATRANS) | Efficiency | 6.020 ± 0.101 | 5.870 ± 0.193 | 0.023 | 0.077 |
| Modularity | 0.079 ± 0.015 | 0.096 ± 0.027 | 0.514 | 1.000 | |
| Clustering coefficient | 5.780 ± 0.111 | 5.650 ± 0.175 | 0.090 | 0.261 | |
| Mean strength | 411.000 ± 14.600 | 390.000 ± 28.800 | 0.023 | 0.070 | |
| NOS | Efficiency | 2109 ± 83.500 | 2120 ± 104.000 | 0.494 | / |
| Modularity | 0.360 ± 0.025 | 0.382 ± 0.039 | 0.867 | / | |
| Clustering coefficient | 197 ± 7.880 | 203 ± 9.630 | 0.243 | / | |
| Mean strength | 43,658 ± 1346 | 43,094 ± 1920 | 0.867 | / |
Results of group comparison performed with robust linear model accounting for gender, age, and density as covariates. To account for multiple comparison, we applied Holm post hoc correction (1) for each network metrics of each microstructural map (adjusted p value metrics) and (ii) for each network metrics extracted from all the microstructural maps of each diffusion-based model (adjusted p value model). The statistically significant results are highlighted in bold.
EXTRAMD, extraneurite mean diffusivity; EXTRATRANS, extraneurite transverse diffusivity; FA, fractional anisotropy; ICVF, intraneurite volume fraction; INTRA, neurite volume fraction; ISOVF, isotropic volume fraction; MD, mean diffusivity; MS, multiple sclerosis; NOS, number of streamlines; RD, radial diffusivity.