Data Synthesis Improves 3D Myotube Instance Segmentation

2026-04-16

AI Summary

This study addresses the challenge of generalizing biomedical segmentation models to 3D myotube instance segmentation due to the scarcity of large-scale annotated data. The authors propose a geometry-driven synthetic myotube generation method that integrates biophysical priors—including polynomial centerlines, variable radii, branching structures, and ellipsoidal end caps—and further enhances realism through the incorporation of authentic noise, optical artifacts, and CycleGAN-based domain adaptation. A lightweight 3D U-Net trained exclusively on this synthetic data, combined with self-supervised encoder pretraining, achieves high-performance myotube instance segmentation without any real annotations. Evaluated on real microscopy images, the method attains an average Intersection-over-Union-based Panoptic Quality (IPQ) of 0.22, substantially outperforming three state-of-the-art zero-shot segmentation approaches.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Web Mining and Content Analysis: Web data generation and simulation
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Myotubes are multinucleated muscle fibers serving as key model systems for studying muscle physiology, disease mechanisms, and drug responses. Mechanistic studies and drug screening thereby rely on quantitative morphological readouts such as diameter, length, and branching degree, which in turn require precise three-dimensional instance segmentation. Yet established pretrained biomedical segmentation models fail to generalize to this domain due to the absence of large annotated myotube datasets. We introduce a geometry-driven synthesis pipeline that models individual myotubes via polynomial centerlines, locally varying radii, branching structures, and ellipsoidal end caps derived from real microscopy observations. Synthetic volumes are rendered with realistic noise, optical artifacts, and CycleGAN-based Domain Adaptation (DA). A compact 3D U-Net with self-supervised encoder pretraining, trained exclusively on synthetic data, achieves a mean IPQ of 0.22 on real data, significantly outperforming three established zero-shot segmentation models, demonstrating that biophysics-driven synthesis enables effective instance segmentation in annotation-scarce biomedical domains.

Problem

• myotube
• 3D instance segmentation
• data scarcity
• biomedical image analysis
• annotation

Innovation

• data synthesis
• 3D instance segmentation
• geometry-driven modeling
• domain adaptation
• self-supervised pretraining