Transductive image segmentation: Self-training and effect of uncertainty estimation
This work addresses the need for optimal predictions in population studies for medical image analysis, though it is incremental as it builds on existing self-training methods.
The paper tackled the problem of improving segmentation predictions on a specific unlabeled dataset in medical imaging, rather than generalizing to unseen data, by exploring transductive self-training with uncertainty estimation, showing promising results in multi-class brain lesion segmentation on MRI data.
Semi-supervised learning (SSL) uses unlabeled data during training to learn better models. Previous studies on SSL for medical image segmentation focused mostly on improving model generalization to unseen data. In some applications, however, our primary interest is not generalization but to obtain optimal predictions on a specific unlabeled database that is fully available during model development. Examples include population studies for extracting imaging phenotypes. This work investigates an often overlooked aspect of SSL, transduction. It focuses on the quality of predictions made on the unlabeled data of interest when they are included for optimization during training, rather than improving generalization. We focus on the self-training framework and explore its potential for transduction. We analyze it through the lens of Information Gain and reveal that learning benefits from the use of calibrated or under-confident models. Our extensive experiments on a large MRI database for multi-class segmentation of traumatic brain lesions shows promising results when comparing transductive with inductive predictions. We believe this study will inspire further research on transductive learning, a well-suited paradigm for medical image analysis.