Table 6.
Classification Metrics | cMRI | Multi-Shell dMRI | cMRI and Multi-Shell dMRI |
---|---|---|---|
Overall performance | |||
Accuracy | 57 ± 8% | 56 ± 7% | 60 ± 5% |
Precision 1 | 0.60 ± 0.14 | 0.59 ± 0.07 | 0.65 ± 0.05 |
MCC | 0.37 ± 0.14 | 0.35 ± 0.10 | 0.43 ± 0.06 |
Individual group sensitivity | |||
IDH-mutant 1p/19q-codeleted | 60 ± 19% | 53 ± 18% | 63 ± 18% |
IDH-mutant 1p/19q-uncodeleted | 37 ± 22% | 43 ± 9% | 43 ± 22% |
IDH wildtype | 73 ± 19% | 70 ± 7% | 73 ± 15% |
Classification performance of ResNet on the test sets with different MRI inputs to diagnose glioma molecular subtype according to WHO CNS5 on the whole cohort of 146 patients. Data are reported as average ± standard deviation over the five test sets in the outer loop. 1 Computed as the macro-average of the precisions for the three groups. WHO CNS5 = fifth edition of the WHO classification of the central nervous system tumours; IDH = isocitrate dehydrogenase; cMRI = conventional MRI; dMRI = diffusion MRI; MCC = Matthews correlation coefficient.