Uncertainty-Aware Temporal Self-Learning (UATS): Semi-Supervised Learning for Segmentation of Prostate Zones and Beyond
This work addresses the problem of insufficient labeled data for medical image segmentation, particularly for fine-grained prostate structures, by applying a semi-supervised learning technique to reduce manual labeling costs, though it is incremental as it builds on existing SSL methods.
The paper tackled the challenging problem of fine-grained segmentation of prostate zones and other structures, which had not reached human-level performance, by proposing an uncertainty-aware temporal self-learning (UATS) method for semi-supervised learning. The result was a significant improvement over supervised baselines, achieving Dice coefficients of up to 78.9%, 87.3%, 75.3%, and 50.6% for different zones, with performance in the range of human inter-rater levels.
Various convolutional neural network (CNN) based concepts have been introduced for the prostate's automatic segmentation and its coarse subdivision into transition zone (TZ) and peripheral zone (PZ). However, when targeting a fine-grained segmentation of TZ, PZ, distal prostatic urethra (DPU) and the anterior fibromuscular stroma (AFS), the task becomes more challenging and has not yet been solved at the level of human performance. One reason might be the insufficient amount of labeled data for supervised training. Therefore, we propose to apply a semi-supervised learning (SSL) technique named uncertainty-aware temporal self-learning (UATS) to overcome the expensive and time-consuming manual ground truth labeling. We combine the SSL techniques temporal ensembling and uncertainty-guided self-learning to benefit from unlabeled images, which are often readily available. Our method significantly outperforms the supervised baseline and obtained a Dice coefficient (DC) of up to 78.9% , 87.3%, 75.3%, 50.6% for TZ, PZ, DPU and AFS, respectively. The obtained results are in the range of human inter-rater performance for all structures. Moreover, we investigate the method's robustness against noise and demonstrate the generalization capability for varying ratios of labeled data and on other challenging tasks, namely the hippocampus and skin lesion segmentation. UATS achieved superiority segmentation quality compared to the supervised baseline, particularly for minimal amounts of labeled data.