Table 3 Top 10 most discriminating categorized features.

From: Multi-feature fusion RFE random forest for schizophrenia classification and treatment response prediction

 

Brain Area 1

Network

Brain Area 2

Network

Feature weight

1

PCUN _L

Default mode

PCUN _R

Default mode

0.1133

2

DCG _L

Subcortical

DCG _R

Subcortical

0.1004

3

CAL _L

Visual

LING _L

Visual

0.0602

4

CUN _L

Visual

CUN _R

Visual

0.0548

5

CAL _L

Visual

CAL _R

Visual

0.0540

6

SMA _L

Attention

SMA _R

Sensorimotor

0.0533

7

SFGmed _L

Default mode

SFGmed _R

Default mode

0.0454

8

ACG _L

Default mode

ACG _R

Default mode

0.0399

9

MTG _L

Default mode

MTG _R

Default mode

0.0368

10

SPG _L

Sensorimotor

IPL _L

Attention

0.0318

  1. Feature Weight: the contribution of features in classification or regression task results; _L: left hemisphere; _R: right hemisphere; PCUN: Precuneus; DCG: Median cingulate and paracingulate gyri; CAL: Calcarine; LING: Lingual; CUN: Cuneus; SMA: Supplementary motor area; SFGmed: medial region of the superior frontal gyrus; ACG: Anterior cingulate and paracingulate gyri; MTG: Middle temporal gyrus; SPG: Superior parietal gyrus; IPL: inferior parietal marginal angular gyrus.
Back to article page