Skip to main content

Table 6 Machine learning results for local clinical data

From: Comparison of machine learning and semi-quantification algorithms for (I123)FP-CIT classification: the beginning of the end for semi-quantification?

Method number

Feature

No. PCs

Kernel

Mean accuracy

SD

Sensitivity

SD

Specificity

SD

ML 1

PCs

3

Linear

0.91

0.05

0.93

0.05

0.88

0.10

ML 2

PCs

5

Linear

0.92

0.05

0.94

0.06

0.88

0.10

ML 3

PCs

10

Linear

0.91

0.05

0.93

0.06

0.86

0.10

ML 4

PCs

15

Linear

0.89

0.05

0.92

0.06

0.83

0.11

ML 5

PCs

20

Linear

0.89

0.05

0.92

0.07

0.83

0.12

ML 6

PCs

3

RBF

0.91

0.05

0.91

0.07

0.89

0.09

ML 7

PCs

5

RBF

0.91

0.06

0.92

0.06

0.89

0.10

ML 8

PCs

10

RBF

0.90

0.05

0.91

0.07

0.88

0.09

ML 9

PCs

15

RBF

0.89

0.05

0.91

0.07

0.87

0.10

ML 10

PCs

20

RBF

0.90

0.05

0.90

0.07

0.89

0.10

ML 11

Voxels

 

Linear

0.88

0.05

0.91

0.06

0.84

0.11

ML 12

SBRs

 

Linear

0.89

0.05

0.92

0.06

0.82

0.10

ML 13

SBRs

 

RBF

0.89

0.06

0.91

0.07

0.85

0.10