TY - GEN
T1 - A Deep Active Learning Framework with Information Guided Label Generation for Medical Image Segmentation
AU - Huang, Kuan
AU - Huang, Jianhua
AU - Wang, Weichen
AU - Xu, Meng
AU - Liu, Feifei
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - Acquiring medical images and their segmentation labels is often time-consuming and labor-intensive. Compared to natural images, medical images need to be screened and annotated by professional doctors in segmentation tasks, especially those containing multiple organ tissues. To reduce the workload of doctors, we propose a deep active learning-based framework for medical image segmentation. Instead of all, the proposed framework can select an optimal number of medical images for label generation by doctors. The selected images are enough to train a good segmentation model because they are diversified, informational, and unique to represent the whole dataset. The proposed method consists of three steps: (1) Using an auto-encoder to extract features from unlabeled images and applying feature clustering and image entropy to select an initial subset of images for segmentation label generation; (2) Employing the mean-teacher method to train a segmentation model in a semi-supervised manner; (3) Updating the labeled subset and reducing redundancy by a deduplication uncertainty query strategy. Extensive experiments are conducted on a breast ultrasound dataset and a knee cartilage ultrasound dataset to evaluate the performance of the proposed method. Experimental results prove that our deep active learning framework can significantly reduce the number of labeled samples while achieving comparable segmentation results to fully-labeled supervision.
AB - Acquiring medical images and their segmentation labels is often time-consuming and labor-intensive. Compared to natural images, medical images need to be screened and annotated by professional doctors in segmentation tasks, especially those containing multiple organ tissues. To reduce the workload of doctors, we propose a deep active learning-based framework for medical image segmentation. Instead of all, the proposed framework can select an optimal number of medical images for label generation by doctors. The selected images are enough to train a good segmentation model because they are diversified, informational, and unique to represent the whole dataset. The proposed method consists of three steps: (1) Using an auto-encoder to extract features from unlabeled images and applying feature clustering and image entropy to select an initial subset of images for segmentation label generation; (2) Employing the mean-teacher method to train a segmentation model in a semi-supervised manner; (3) Updating the labeled subset and reducing redundancy by a deduplication uncertainty query strategy. Extensive experiments are conducted on a breast ultrasound dataset and a knee cartilage ultrasound dataset to evaluate the performance of the proposed method. Experimental results prove that our deep active learning framework can significantly reduce the number of labeled samples while achieving comparable segmentation results to fully-labeled supervision.
KW - deep active learning
KW - label generation workflow
KW - medical image segmentation
UR - http://www.scopus.com/inward/record.url?scp=85146725299&partnerID=8YFLogxK
U2 - 10.1109/BIBM55620.2022.9995046
DO - 10.1109/BIBM55620.2022.9995046
M3 - Conference contribution
AN - SCOPUS:85146725299
T3 - Proceedings - 2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2022
SP - 1562
EP - 1567
BT - Proceedings - 2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2022
A2 - Adjeroh, Donald
A2 - Long, Qi
A2 - Shi, Xinghua
A2 - Guo, Fei
A2 - Hu, Xiaohua
A2 - Aluru, Srinivas
A2 - Narasimhan, Giri
A2 - Wang, Jianxin
A2 - Kang, Mingon
A2 - Mondal, Ananda M.
A2 - Liu, Jin
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2022
Y2 - 6 December 2022 through 8 December 2022
ER -