Synaptic protein segmentation - Zooniverse - STED-FM

This is the dataset associated with the Synapses segmentation in "A foundation model for STED microscopy enabling multi-task analysis, representation-based discovery, and interactive microscopy".

The dataset was annotated in a community-driven approach using the Zooniverse project. See the website for more information about the project! The images associated with this dataset were acquired in the following paper.

Wiesner, T., Bilodeau, A., Bernatchez, R., Raulier, B., De Koninck, P., & Lavoie-Cardinal, F. (2020). Activity-dependent remodelling of synaptic protein organization revealed by high throughput analysis of STED nanoscopy images. Frontiers in Neural Circuits, 14. https://doi.org/10.3389/fncir.2020.00057

The information related to how the images were acquired are provided in this paper.

The dataset is already split into training, validation, and testing datasets and provided each in their respective `.hdf5` files. We recommend using the code provided in the GitHub repository of the paper. We provide a simple python script (`synaptic-protein-segmentation.py`) for loading the datasets.

本数据集关联《用于受激发射损耗（STED）显微镜的基础模型，支持多任务分析、基于表征的发现与交互式显微成像》一文中的突触分割任务。

本数据集依托Zooniverse项目，采用社区共建模式完成标注。欲了解该项目详情，请访问其官方网站。本数据集关联的成像数据采集自下述论文：

Wiesner T, Bilodeau A, Bernatchez R, Raulier B, De Koninck P, Lavoie-Cardinal F. 高通量分析STED超分辨显微镜图像揭示突触蛋白组织的活动依赖性重塑. Frontiers in Neural Circuits, 2020, 14. https://doi.org/10.3389/fncir.2020.00057

该论文中详细说明了图像的采集细节。

本数据集已划分为训练集、验证集与测试集，分别存储于对应的`.hdf5`格式文件中。我们建议使用该论文GitHub仓库中提供的代码开展相关操作，同时附带了用于加载数据集的简易Python脚本`synaptic-protein-segmentation.py`。