TABLE 2.
Summary of the PMA regression results. ENA50 column shows the results of our proposed method with the parcellation nodes defined with the ENA50 atlas. The other four columns present the results of four different ways of implementing connectome‐level averaging, including length‐preserved averaging (column 2), simple average (column 3), consensus‐based uniform thresholding with (column 4), and consensus‐based uniform thresholding with a threshold (column 5).
| ENA50 | Length‐preserved | Simple average | Consensus | Consensus | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| p‐value | R | p‐value | R | p‐value | R | p‐value | R | p‐value | R | ||
| FA | GE | <.0001 | .9854 | <.0001 | .9777 | <.0001 | .9905 | <.0001 | .9671 | <.0001 | .9766 |
| LE | <.0001 | .9858 | <.0001 | .9699 | <.0001 | .9907 | <.0001 | .9481 | <.0001 | .9872 | |
| CPL | <.0001 | −.9728 | <.0001 | −.9603 | <.0001 | −.9775 | <.0001 | −.9586 | <.0001 | −.9592 | |
| CC | <.0001 | .9772 | <.0001 | .9753 | <.0001 | .9894 | <.0001 | .9517 | <.0001 | .9746 | |
| SWI | <.0001 | / | <.0001 | / | .1098 | / | <.0001 | 1.0000 | / | ||
| NDI | GE | <.0001 | .9813 | <.0001 | .9829 | <.0001 | .9926 | <.0001 | .9756 | <.0001 | .9809 |
| LE | <.0001 | .9794 | <.0001 | .9770 | <.0001 | .9920 | <.0001 | .9639 | <.0001 | .9692 | |
| CPL | <.0001 | −.9730 | <.0001 | −.9700 | <.0001 | −.9829 | <.0001 | −.9620 | <.0001 | −.9638 | |
| CC | <.0001 | .9750 | <.0001 | .9798 | <.0001 | .9914 | <.0001 | .9661 | <.0001 | .9648 | |
| SW | <.0001 | / | .0074 | / | .8537 | / | .0293 | / | 1.0000 | / | |
| 1‐ODI | GE | .0111 | .7008 | .4092 | .2628 | .7368 | −.1086 | .2748 | .3431 | .7386 | .1079 |
| LE | .0144 | .6826 | .6256 | .1572 | .4145 | −.2600 | .4607 | .2358 | .9868 | .0054 | |
| CPL | .0140 | −.6850 | .3283 | −.3091 | .5306 | −.2012 | .3217 | −.3131 | .3532 | −.2942 | |
| CC | .0353 | .6098 | .8440 | .0638 | .6820 | .1323 | .6498 | .1464 | .5419 | −.1958 | |
| SW | .0020 | / | <.0001 | / | .0012 | / | <.0001 | / | .0005 | / | |