Table 1.
Accuracy of EpICC for classification of cancer types and sub-types in comparison to published methods.
| Study | Classification accuracy (%) | ||||
|---|---|---|---|---|---|
| Cancer types | LGG subtypes | BRCA subtypes | ESCA subtypes | THCA subtypes | |
| Lyu and Haque29 | 95.59% (33) | NA | NA | NA | NA |
| Kim et al.31 | 91.74% (21) | NA | NA | NA | NA |
| Xiao et al.25 | 96%-99% (3) | NA | NA | NA | NA |
| Ramirez et al.49 | 94.70% (33) | NA | NA | NA | NA |
| Sun et al.48 | 97.47% (12) | NA | NA | NA | NA |
| Pei et al.50 | NA | 63.90 (3) | NA | NA | NA |
| Couture et al.51 | NA | NA | 94 (2) | NA | NA |
| EpICC | 97.83% (31) | 81.31 (3) | 94.98 (2) | 97.5 (3) | 95.24 (2) |
The number within the brackets show the number of cancer types or subtypes for which classifications were done.