ANNOTATION-EFFICIENT 3D U-NETS FOR BRAIN PLASTICITY NETWORK MAPPING

A fundamental challenge in machine learning-based segmentation of large-scale brain microscopy images is the time and domain expertise required by humans to generate ground truth for model training. Weakly supervised and semi-supervised approaches can greatly reduce the burden of human annotation. Here we present a study of three-dimensional U-Nets with varying levels of supervision to perform neuronal nuclei segmentation in light-sheet microscopy volumes. We leverage automated blob detection with classical algorithms to generate noisy labels on a large volume, and our experiments show that weak supervision, with or without additional fine-tuning, can outperform resource-limited fully supervised learning. These methods are extended to analyze coincidence between multiple fluorescent stains in cleared brain tissue. This is an initial step towards automated whole-brain analysis of plasticity-related gene expression.

Automated summary

The study examines the use of three-dimensional U-Nets for neuronal nuclei segmentation, demonstrating that weakly supervised and semi-supervised methods can reduce the burden of human annotation. By leveraging automated blob detection with classical algorithms to generate noisy labels, the experiments show that weak supervision can outperform fully supervised learning in resource-limited settings. These methods are also extended to analyze the coincidence of multiple fluorescent stains in cleared brain tissue, representing a step towards automated whole-brain analysis of plasticity-related gene expression.