Research Article
Tongue Image Database Construction Based on the Expert Opinions: Assessment for Individual Agreement and Methods for Expert Selection
Table 4
Optimizing results of interrater agreement for all the tongue features which needed optimization.
| Optimized tongue features | recognition times | rest of the experts | Inter-rater agreement | Before optimization | After optimization |
| Tongue body | Light red | 3 | 19 | 0.3431 | 0.6450 | moderate texture | 8 | 16 | 0.0948 | 0.6333 | tender | 2 | 41 | 0.5349 | 0.6535 | Enlarged | 3 | 26 | 0.3345 | 0.6404 | Moderate shape | 5 | 10 | 0.1477 | 0.6310 | Teeth-print | 2 | 16 | 0.4580 | 0.6363 | crack | 2 | 21 | 0.4972 | 0.6960 |
| Tongue fur | White fur | 2 | 35 | 0.4902 | 0.6907 | Thin fur | 3 | 24 | 0.2769 | 0.4601 | Thick fur | 3 | 4 | 0.3147 | 0.6406 | Moist fur | 5 | 22 | 0.2706 | 0.6989 | Greasy fur | 4 | 24 | 0.2213 | 0.6200 |
|
|