Test-Time Augmentation-Based Active Learning and Self-training for Label-Efficient Segmentation

Bella Specktor-Fadida*, Anna Levchakov, Dana Schonberger, Liat Ben-Sira, Dafna Ben Bashat, Leo Joskowicz

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations


Deep learning techniques depend on large datasets whose annotation is time-consuming. To reduce annotation burden, the self-training (ST) and active-learning (AL) methods have been developed as well as methods that combine them in an iterative fashion. However, it remains unclear when each method is the most useful, and when it is advantageous to combine them. In this paper, we propose a new method that combines ST with AL using Test-Time Augmentations (TTA). First, TTA is performed on an initial teacher network. Then, cases for annotation are selected based on the lowest estimated Dice score. Cases with high estimated scores are used as soft pseudo-labels for ST. The selected annotated cases are trained with existing annotated cases and ST cases with border slices annotations. We demonstrate the method on MRI fetal body and placenta segmentation tasks with different data variability characteristics. Our results indicate that ST is highly effective for both tasks, boosting performance for in-distribution (ID) and out-of-distribution (OOD) data. However, while self-training improved the performance of single-sequence fetal body segmentation when combined with AL, it slightly deteriorated performance of multi-sequence placenta segmentation on ID data. AL was helpful for the high variability placenta data, but did not improve upon random selection for the single-sequence body data. For fetal body segmentation sequence transfer, combining AL with ST following ST iteration yielded a Dice of 0.961 with only 6 original scans and 2 new sequence scans. Results using only 15 high-variability placenta cases were similar to those using 50 cases. Code is available at: https://github.com/Bella31/TTA-quality-estimation-ST-AL.

Original languageAmerican English
Title of host publicationMedical Image Learning with Limited and Noisy Data - 2nd International Workshop, MILLanD 2023, Held in Conjunction with MICCAI 2023, Proceedings
EditorsZhiyun Xue, Sameer Antani, Ghada Zamzmi, Feng Yang, Sivaramakrishnan Rajaraman, Zhaohui Liang, Sharon Xiaolei Huang, Marius George Linguraru
PublisherSpringer Science and Business Media Deutschland GmbH
Number of pages11
ISBN (Print)9783031471964
StatePublished - 2023
Event2nd Workshop on Medical Image Learning with Noisy and Limited Data, MILLanD 2023 - Vancouver, Canada
Duration: 8 Oct 20238 Oct 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14307 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference2nd Workshop on Medical Image Learning with Noisy and Limited Data, MILLanD 2023

Bibliographical note

Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.


  • Active learning
  • Fetal MRI
  • Self-training
  • Test-time augmentation


Dive into the research topics of 'Test-Time Augmentation-Based Active Learning and Self-training for Label-Efficient Segmentation'. Together they form a unique fingerprint.

Cite this