Table 2 The cross-validated diagnostic accuracy of the LDA for the three cerebellar tumor types.

From: Tissue metabolite profiles for the characterisation of paediatric cerebellar tumours

Classification accuracy of linear discriminant analysis

 

Ependymoma

Medulloblastoma

Pilocytic astrocytoma

% correct

Ependymoma

17

0

1

94.4

Medulloblastoma

1

34

1

94.4

Pilocytic astrocytoma

1

1

22

91.7

Rapid intraoperative histopathological diagnostic accuracy

 

Number of concordant diagnoses

Number of partially concordant diagnoses

Number of discordant diagnoses

% concordant

Ependymoma

10

4

0

71.4

Medulloblastoma

28

1

2

90.3

Pilocytic astrocytoma

16

0

2

88.9

  1. The linear discriminant classifier achieved accuracies greater than 90% for all three tumor types. Furthermore, for all tumor types, the classification accuracy was greater than the rapid intraoperative diagnostic examination; however, this appears to be due to partial concordance as opposed to an incorrect diagnosis.
Back to article page