Figure 2 | Scientific Reports

Figure 2

From: Training calibration-based counterfactual explainers for deep learning models in medical image analysis

Figure 2Figure 2

(a–g) Diagnosis-based counterfactual explanations generated using TraCE by progressively introducing relevant patterns into different query images (first image in each row) of healthy subjects to increase the likelihood of being assigned to the abnormal group (\(P(state = {{abnormal}})\)). (h–n) Diagnosis-based counterfactual explanations generated using TraCE by progressively introducing relevant patterns into different query images (first image in each row) of healthy subjects to increase the likelihood of being assigned to the abnormal group (\(P(state = {{abnormal}})\)).

Back to article page