Table 3:

The performance of feature combinations with the SVM classifier in differentiating long- and short-term survival time

Feature SetNFaSVM
SensitivitySpecificityAccuracyAUC
CMb260.79100.66670.72930.7786
RLM220.73130.42420.57890.5880
Histogram80.68660.25760.47370.4326
CM + RLMb480.77610.77270.77440.8053
CM + histogram340.77610.68180.72930.7782
RLM + histogram300.80600.37880.59400.6223
CM + RLM + histogramc560.77610.78790.78200.8057

  • Note:—AUC indicates area under the curve.

  • ↵a The number of features in each subset.

  • ↵b The best classification when the feature type is used alone or pair-combined.

  • ↵c The best classification result when all features are used.