DOMINO: Domain-aware Loss for Deep Learning Calibration
This addresses the critical need for reliable model predictions in high-risk medical applications, though it appears incremental as it builds on existing calibration methods with a domain-specific twist.
The paper tackles the problem of model calibration in deep learning for medical imaging by proposing a domain-aware loss function that applies class-wise penalties based on class similarity, resulting in improved calibration and reduced risky errors.
Deep learning has achieved the state-of-the-art performance across medical imaging tasks; however, model calibration is often not considered. Uncalibrated models are potentially dangerous in high-risk applications since the user does not know when they will fail. Therefore, this paper proposes a novel domain-aware loss function to calibrate deep learning models. The proposed loss function applies a class-wise penalty based on the similarity between classes within a given target domain. Thus, the approach improves the calibration while also ensuring that the model makes less risky errors even when incorrect. The code for this software is available at https://github.com/lab-smile/DOMINO.