Table 2 Comparison of performance of cluster splitting methods on BBBC0004 dataset (None—images without applied cluster splitting method)
From: Clustered nuclei splitting based on recurrent distance transform in digital pathology images
| Â | Â | TPR | PPV | F1 | MDR | FDR | USR | OSR |
|---|---|---|---|---|---|---|---|---|
ov00 | None | 0.968 | 1.000 | 0.984 | 0.00 | 0.00 | 0.03 | 0.00 |
| Â | Watershed | 0.978 | 1.000 | 0.989 | 0.00 | 0.00 | 0.02 | 0.00 |
| Â | Proposed method | 0.989 | 1.000 | 0.995 | 0.00 | 0.00 | 0.01 | 0.00 |
| Â | H-minima | 0.989 | 1.000 | 0.995 | 0.00 | 0.00 | 0.01 | 0.00 |
| Â | Huang | 0.989 | 0.999 | 0.994 | 0.00 | 0.00 | 0.01 | 0.00 |
| Â | Kong | 0.989 | 0.999 | 0.994 | 0.00 | 0.00 | 0.01 | 0.00 |
| Â | Mouelhi | 0.968 | 1.000 | 0.984 | 0.00 | 0.00 | 0.03 | 0.00 |
ov15 | None | 0.793 | 0.997 | 0.884 | 0.00 | 0.00 | 0.21 | 0.00 |
| Â | Watershed | 0.811 | 0.997 | 0.894 | 0.00 | 0.00 | 0.19 | 0.00 |
|  | Proposed method | 0.970 †| 0.970 | 0.970 †| 0.01 | 0.00 | 0.02 | 0.02 |
| Â | H-minima | 0.934 | 0.991 | 0.961 | 0.00 | 0.00 | 0.06 | 0.01 |
| Â | Huang | 0.963 | 0.955 | 0.959 | 0.02 | 0.00 | 0.01 | 0.04 |
| Â | Kong | 0.955 | 0.972 | 0.963 | 0.01 | 0.00 | 0.04 | 0.03 |
| Â | Mouelhi | 0.826 | 0.995 | 0.902 | 0.00 | 0.00 | 0.17 | 0.00 |
ov30 | None | 0.723 | 1.000 | 0.839 | 0.00 | 0.00 | 0.28 | 0.00 |
| Â | Watershed | 0.739 | 1.000 | 0.850 | 0.00 | 0.00 | 0.26 | 0.00 |
|  | Proposed method | 0.959 †| 0.986 | 0.972 †| 0.01 | 0.00 | 0.03 | 0.01 |
| Â | H-minima | 0.890 | 0.999 | 0.941 | 0.00 | 0.00 | 0.11 | 0.00 |
| Â | Huang | 0.950 | 0.963 | 0.956 | 0.04 | 0.01 | 0.02 | 0.03 |
| Â | Kong | 0.926 | 0.982 | 0.953 | 0.00 | 0.00 | 0.07 | 0.02 |
| Â | Mouelhi | 0.774 | 0.996 | 0.871 | 0.00 | 0.00 | 0.22 | 0.00 |
ov45 | None | 0.700 | 1.000 | 0.824 | 0.00 | 0.00 | 0.30 | 0.00 |
| Â | Watershed | 0.716 | 1.000 | 0.834 | 0.00 | 0.00 | 0.28 | 0.00 |
|  | Proposed method | 0.942 | 0.993 | 0.967 †| 0.01 | 0.00 | 0.05 | 0.01 |
| Â | H-minima | 0.873 | 0.999 | 0.932 | 0.00 | 0.00 | 0.13 | 0.00 |
|  | Huang | 0.946 †| 0.981 | 0.963 | 0.03 | 0.00 | 0.02 | 0.02 |
| Â | Kong | 0.930 | 0.992 | 0.960 | 0.00 | 0.00 | 0.07 | 0.01 |
| Â | Mouelhi | 0.756 | 0.999 | 0.861 | 0.00 | 0.00 | 0.24 | 0.00 |
ov60 | None | 0.610 | 1.000 | 0.757 | 0.00 | 0.00 | 0.39 | 0.00 |
| Â | Watershed | 0.626 | 1.000 | 0.770 | 0.00 | 0.00 | 0.37 | 0.00 |
|  | Proposed method | 0.909 | 0.983 | 0.945 †| 0.01 | 0.00 | 0.08 | 0.02 |
| Â | H-minima | 0.809 | 0.993 | 0.891 | 0.00 | 0.00 | 0.19 | 0.01 |
|  | Huang | 0.921 †| 0.973 | 0.945 †| 0.05 | 0.00 | 0.03 | 0.03 |
| Â | Kong | 0.893 | 0.964 | 0.927 | 0.01 | 0.00 | 0.09 | 0.04 |
| Â | Mouelhi | 0.696 | 0.993 | 0.818 | 0.00 | 0.00 | 0.30 | 0.01 |