Fig. 9.

Boxplot comparison of aggregated Dice Similarity Coefficient (DSCagg) scores across Task 1 (pre-radiotherapy) and Task 2 (mid-radiotherapy) for three metrics: DSCagg mean, DSCagg primary gross tumor volume (GTVp), and DSCagg metastatic lymph nodes (GTVn). Each box represents the interquartile range, with the horizontal line indicating the median score. Outliers are shown as individual points outside the whiskers. Task 1 is represented in red, and Task 2 in blue. Scatter symbols indicate the nnU-Net baseline (triangle), and interobserver variability (IOV, inverted triangle).