Table 2 Confusion matrix between pathologist-based scoring and automated scoring for the seventy-one cases in the cohort.
From: Relevance of deep learning to facilitate the diagnosis of HER2 status in breast cancer
Pathologist-based scores | ||||
|---|---|---|---|---|
Negative | Equivocal | Positive | ||
Deep learning based scores | Negative | 41 | 3 | 0 |
Equivocal | 2 | 8 | 1 | |
Positive | 0 | 6 | 10 | |
Overall agreement = 0.83 (95CI: 0.74–0.92) | ||||
Cohen’s κ = 0.69 (95CI: 0.55–0.84) | ||||
Kendall’s τ = 0.84 (95CI: 0.75–0.93) | ||||
