Skip to main content

Table 1 Comparing results for the validation of SC, using PCA prior to LDA.

From: SpectraClassifier 1.0: a user friendly, automated MRS-based classifier-development system

  Long TE - [20] Long TE - SC Short TE - [21] Short TE - SC
Classes AUC ± SE AUC ± SE AUC ± SE AUC ± SE
1 vs. 2 0.953 ± 0.031 (8) 0.977 ± 0.016 (8) 0.956 ± 0.028 (4) 0.923 ± 0.028 (4)
1 vs. 3 0.593 ± 0.104 (6) 0.757 ± 0.054 (6) 0.591 ± 0.097 (4) 0.688 ± 0.056 (4)
1 vs. 4 0.918 ± 0.063 (7) 0.941 ± 0.025 (7) 0.966 ± 0.029 (3) 0.962 ± 0.019 (3)
2 vs. 3 0.961 ± 0.038 (5) 0.970 ± 0.028 (5) 0.954 ± 0.044 (4) 0.972 ± 0.021 (4)
2 vs. 4 0.931 ± 0.073 (10) 0.999 ± 0.003 (10) 0.997 ± 0.009 (11) 1.000 ± 0.000 (11)
3 vs. 4 0.961 ± 0.053 (4) 0.995 ± 0.010 (4) 0.986 ± 0.025 (2) 0.979 ± 0.025 (2)
  1. In this example, multiple binary classifiers were developed for long and short TE of SV MRS data from INTERPRET [5]. In [20, 21] the PCA covered the 75% of the variance of the dataset. For SC, a variance of 80% has been covered. As performance measure, the AUC and its standard error (SE) were used. The number between brackets refers to the number of principal components used. The classes are: 1) glioblastomas, 2) meningiomas, 3) metastases and 4) astrocytomas grade II.