Mutual Evidential Deep Learning for Semi-supervised Medical Image Segmentation

Existing semi-supervised medical segmentation co-learning frameworks have realized that model performance can be diminished by the biases in model recognition caused by low-quality pseudo-labels. Due to the averaging nature of their pseudo-label integration strategy, they fail to explore the reliabi...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Proceedings (IEEE International Conference on Bioinformatics and Biomedicine) s. 2010 - 2017
Hlavní autoři: He, Yuanpeng, Bi, Yali, Li, Lijian, Pun, Chi-Man, Jiao, Wenpin, Jin, Zhi
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 03.12.2024
Témata:
Biomedical imaging
Class-aware evidential fusion
Data models
Deep learning
Heterogeneous networks
Image segmentation
Mutual evidential deep learning
Predictive models
Reliability
Semi-supervised medical segmentation
Semisupervised learning
Training
Uncertainty
ISSN:2156-1133
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Existing semi-supervised medical segmentation co-learning frameworks have realized that model performance can be diminished by the biases in model recognition caused by low-quality pseudo-labels. Due to the averaging nature of their pseudo-label integration strategy, they fail to explore the reliability of pseudo-labels from different sources. In this paper, we propose a mutual evidential deep learning (MEDL) framework that offers a potentially viable solution for pseudo-label generation in semi-supervised learning from two perspectives. First, we introduce networks with different architectures to generate complementary evidence for unlabeled samples and adopt an improved class-aware evidential fusion to guide the confident synthesis of evidential predictions sourced from diverse architectural networks. Second, utilizing the uncertainty in the fused evidence, we design an asymptotic Fisher information-based evidential learning strategy. This strategy enables the model to initially focus on unlabeled samples with more reliable pseudo-labels, gradually shifting attention to samples with lower-quality pseudo-labels while avoiding over-penalization of mislabeled classes in high data uncertainty samples. Additionally, for labeled data, we continue to adopt an uncertainty-driven asymptotic learning strategy, gradually guiding the model to focus on challenging voxels. Extensive experiments on five mainstream datasets have demonstrated that MEDL achieves state-of-the-art performance.
ISSN:2156-1133
DOI:10.1109/BIBM62325.2024.10822008