Boosting Overlapping Organoid Instance Segmentation Using Pseudo-Label Unmixing and Synthesis-Assisted Learning

2026-01-10

arXiv.org

AI Summary

This work addresses the challenge of organoid instance segmentation in highly overlapping microscopy images, where performance is hindered by scarce annotated data and noisy pseudo-labels. To overcome these limitations, we introduce synthetic-assisted semi-supervised learning (SA-SSL) to this task for the first time, proposing a Pseudo-Label Unmixing (PLU) mechanism to disentangle overlapping instances. Our approach further integrates contour-guided organoid synthesis and instance-level augmentation strategies to substantially improve label efficiency. Remarkably, using only 10% of the labeled data, our model achieves performance comparable to fully supervised methods on two organoid datasets, setting a new state-of-the-art in semi-supervised organoid segmentation.

Technology Category

• Machine Learning: Semi-Supervised Learning
• Computer Vision: Segmentation
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labeling

Abstract

Organoids, sophisticated in vitro models of human tissues, are crucial for medical research due to their ability to simulate organ functions and assess drug responses accurately. Accurate organoid instance segmentation is critical for quantifying their dynamic behaviors, yet remains profoundly limited by high-quality annotated datasets and pervasive overlap in microscopy imaging. While semi-supervised learning (SSL) offers a solution to alleviate reliance on scarce labeled data, conventional SSL frameworks suffer from biases induced by noisy pseudo-labels, particularly in overlapping regions. Synthesis-assisted SSL (SA-SSL) has been proposed for mitigating training biases in semi-supervised semantic segmentation. We present the first adaptation of SA-SSL to organoid instance segmentation and reveal that SA-SSL struggles to disentangle intertwined organoids, often misrepresenting overlapping instances as a single entity. To overcome this, we propose Pseudo-Label Unmixing (PLU), which identifies erroneous pseudo-labels for overlapping instances and then regenerates organoid labels through instance decomposition. For image synthesis, we apply a contour-based approach to synthesize organoid instances efficiently, particularly for overlapping cases. Instance-level augmentations (IA) on pseudo-labels before image synthesis further enhances the effect of synthetic data (SD). Rigorous experiments on two organoid datasets demonstrate our method's effectiveness, achieving performance comparable to fully supervised models using only 10% labeled data, and state-of-the-art results. Ablation studies validate the contributions of PLU, contour-based synthesis, and augmentation-aware training. By addressing overlap at both pseudo-label and synthesis levels, our work advances scalable, label-efficient organoid analysis, unlocking new potential for high-throughput applications in precision medicine.

Problem

• organoid instance segmentation
• overlapping instances
• pseudo-label bias
• semi-supervised learning
• label scarcity

Innovation

• Pseudo-Label Unmixing
• Synthesis-Assisted SSL
• Organoid Instance Segmentation
• Contour-Based Synthesis
• Instance-Level Augmentation