Instance-Aware Pseudo-Labeling and Class-Focused Contrastive Learning for Weakly Supervised Domain Adaptive Segmentation of Electron Microscopy

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Not specified in abstract

• Primary Outcome: Segmentation performance improvement over existing UDA and WDA methods

• Nnt Or Sensitivity: Claims to 'significantly narrow the performance gap with supervised upper bound'

• Confidence Interval: Not reported

Clinical Application

High - addresses practical annotation burden in microscopy research Relevant to Australian research institutions with electron microscopy facilities, particularly in neuroscience and cell biology research centres Biological and neuroscience researchers using electron microscopy for mitochondrial analysis

Abstract

Annotation-efficient segmentation of the numerous mitochondria instances from various electron microscopy (EM) images is highly valuable for biological and neuroscience research. Although unsupervised domain adaptation (UDA) methods can help mitigate domain shifts and reduce the high costs of annotating each domain, they typically have relatively low performance in practical applications. Thus, we investigate weakly supervised domain adaptation (WDA) that utilizes additional sparse point labels on the target domain, which require minimal annotation effort and minimal expert knowledge. To take full use of the incomplete and imprecise point annotations, we introduce a multitask learning framework that jointly conducts segmentation and center detection with a novel cross-teaching mechanism and class-focused cross-domain contrastive learning. While leveraging unlabeled image regions is essential, we introduce segmentation self-training with a novel instance-aware pseudo-label (IPL) selection strategy. Unlike existing methods that typically rely on pixel-wise pseudo-label filtering, the IPL semantically selects reliable and diverse pseudo-labels with the help of the detection task. Comprehensive validations and comparisons on challenging datasets demonstrate that our method outperforms existing UDA and WDA methods, significantly narrowing the performance gap with the supervised upper bound. Furthermore, under the UDA setting, our method also achieves substantial improvements over other UDA techniques.

References

1. 1.Xiong, S., Chen, J., Wang, Y., & Peng, J. (2026). Instance-aware pseudo-labeling and class-focused contrastive learning for weakly supervised domain adaptive segmentation of electron microscopy. Neuroinformatics. https://doi.org/10.1109/JPROC.2021.3054390