An Empirical Study on MC Dropout--Based Uncertainty--Error Correlation in 2D Brain Tumor Segmentation
This is an incremental study for medical image segmentation researchers, showing limited utility of MC Dropout for error localization in brain tumor segmentation.
This study investigated the relationship between MC Dropout-based uncertainty and segmentation errors in 2D brain tumor MRI segmentation, finding weak global correlations (r ≈ 0.30–0.38) and negligible boundary correlations (|r| < 0.05).
Accurate brain tumor segmentation from MRI is vital for diagnosis and treatment planning. Although Monte Carlo (MC) Dropout is widely used to estimate model uncertainty, its effectiveness in identifying segmentation errors -- especially near tumor boundaries -- remains unclear. This study empirically examines the relationship between MC Dropout--based uncertainty and segmentation error in 2D brain tumor MRI segmentation using a U-Net trained under four augmentation settings: none, horizontal flip, rotation, and scaling. Uncertainty was computed from 50 stochastic forward passes and correlated with pixel-wise errors using Pearson and Spearman coefficients. Results show weak global correlations ($r \approx 0.30$--$0.38$) and negligible boundary correlations ($|r| < 0.05$). Although differences across augmentations were statistically significant ($p < 0.001$), they lacked practical relevance. These findings suggest that MC Dropout uncertainty provides limited cues for boundary error localization, underscoring the need for alternative or hybrid uncertainty estimation methods in medical image segmentation.