[Retracted] Artificial Intelligence-Guided Subspace Clustering Algorithm for Glioma Images

Table 2

Average clustering results of 10 patients.

Feature extraction method	Dice coefficient (%)	Sensitivity (%)	False negative rate (%)
Gray + Haar wavelet	85.17 ± 1. 64	91. 36 ± 1. 81	17. 55 ± 8. 17
Gray + Haar wavelet + 3D-CNNs	88. 17 ± 1. 21	90. 58 ± 2. 77	16.14 ± 4. 28
Gray + Haar wavelet + 2D-CNNs	85. 93 ± 3. 58	91. 18 ± 3. 48	24. 28 ± 1. 84