Semi-supervised Medical Image Classification with Temporal Knowledge-Aware Regularization
Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Standard
Semi-supervised Medical Image Classification with Temporal Knowledge-Aware Regularization. / Yang, Qiushi; Liu, Xinyu; Chen, Zhen; Ibragimov, Bulat; Yuan, Yixuan.
Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings. red. / Linwei Wang; Qi Dou; P. Thomas Fletcher; Stefanie Speidel; Shuo Li. Springer, 2022. s. 119-129 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bind 13438 LNCS).Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Semi-supervised Medical Image Classification with Temporal Knowledge-Aware Regularization
AU - Yang, Qiushi
AU - Liu, Xinyu
AU - Chen, Zhen
AU - Ibragimov, Bulat
AU - Yuan, Yixuan
N1 - Publisher Copyright: © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - Semi-supervised learning (SSL) for medical image classification has achieved exceptional success on efficiently exploiting knowledge from unlabeled data with limited labeled data. Nevertheless, recent SSL methods suffer from misleading hard-form pseudo labeling, exacerbating the confirmation bias issue due to rough training process. Moreover, the training schemes excessively depend on the quality of generated pseudo labels, which is vulnerable against the inferior ones. In this paper, we propose TEmporal knowledge-Aware Regularization (TEAR) for semi-supervised medical image classification. Instead of using hard pseudo labels to train models roughly, we design Adaptive Pseudo Labeling (AdaPL), a mild learning strategy that relaxes hard pseudo labels to soft-form ones and provides a cautious training. AdaPL is built on a novel theoretically derived loss estimator, which approximates the loss of unlabeled samples according to the temporal information across training iterations, to adaptively relax pseudo labels. To release the excessive dependency of biased pseudo labels, we take advantage of the temporal knowledge and propose Iterative Prototype Harmonizing (IPH) to encourage the model to learn discriminative representations in an unsupervised manner. The core principle of IPH is to maintain the harmonization of clustered prototypes across different iterations. Both AdaPL and IPH can be easily incorporated into prior pseudo labeling-based models to extract features from unlabeled medical data for accurate classification. Extensive experiments on three semi-supervised medical image datasets demonstrate that our method outperforms state-of-the-art approaches. The code is available at https://github.com/CityU-AIM-Group/TEAR.
AB - Semi-supervised learning (SSL) for medical image classification has achieved exceptional success on efficiently exploiting knowledge from unlabeled data with limited labeled data. Nevertheless, recent SSL methods suffer from misleading hard-form pseudo labeling, exacerbating the confirmation bias issue due to rough training process. Moreover, the training schemes excessively depend on the quality of generated pseudo labels, which is vulnerable against the inferior ones. In this paper, we propose TEmporal knowledge-Aware Regularization (TEAR) for semi-supervised medical image classification. Instead of using hard pseudo labels to train models roughly, we design Adaptive Pseudo Labeling (AdaPL), a mild learning strategy that relaxes hard pseudo labels to soft-form ones and provides a cautious training. AdaPL is built on a novel theoretically derived loss estimator, which approximates the loss of unlabeled samples according to the temporal information across training iterations, to adaptively relax pseudo labels. To release the excessive dependency of biased pseudo labels, we take advantage of the temporal knowledge and propose Iterative Prototype Harmonizing (IPH) to encourage the model to learn discriminative representations in an unsupervised manner. The core principle of IPH is to maintain the harmonization of clustered prototypes across different iterations. Both AdaPL and IPH can be easily incorporated into prior pseudo labeling-based models to extract features from unlabeled medical data for accurate classification. Extensive experiments on three semi-supervised medical image datasets demonstrate that our method outperforms state-of-the-art approaches. The code is available at https://github.com/CityU-AIM-Group/TEAR.
KW - Medical image classification
KW - Prototype harmonizing
KW - Pseudo labeling
KW - Semi-supervised learning
KW - Temporal knowledge
U2 - 10.1007/978-3-031-16452-1_12
DO - 10.1007/978-3-031-16452-1_12
M3 - Article in proceedings
AN - SCOPUS:85138992658
SN - 9783031164514
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 119
EP - 129
BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2022 - 25th International Conference, Proceedings
A2 - Wang, Linwei
A2 - Dou, Qi
A2 - Fletcher, P. Thomas
A2 - Speidel, Stefanie
A2 - Li, Shuo
PB - Springer
T2 - 25th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2022
Y2 - 18 September 2022 through 22 September 2022
ER -
ID: 322796989