Table 6. DNN model performance in cancer cell type prediction.
CANCER CELL TYPE | SPECIFICITY | SENSITIVITY (RECALL) | PPV (PRECISION) | NPV |
---|---|---|---|---|
BLADDER | 0.9946 | 0.7778 | 0.8750 | 0.9893 |
BRAIN | 1.0000 | 0.7959 | 1.0000 | 0.9716 |
BREAST | 0.9942 | 0.8958 | 0.9556 | 0.9855 |
COLORECTAL | 0.9942 | 0.9333 | 0.9545 | 0.9914 |
HEAD AND NECK | 1.0000 | 0.8421 | 1.0000 | 0.9833 |
KIDNEY | 0.9889 | 0.9355 | 0.8788 | 0.9944 |
LIVER | 0.9602 | 0.8571 | 0.4444 | 0.9945 |
LUNG | 0.9842 | 0.6267 | 0.9038 | 0.9174 |
PANCREAS | 0.9973 | 0.4167 | 0.9091 | 0.9632 |
SKIN | 0.9826 | 1.0000 | 0.8868 | 1.0000 |
TESTIS | 0.9974 | 1.0000 | 0.6667 | 1.0000 |
OVERALL | 0.9903 | 0.8255 | 0.8613 | 0.9810 |