Multiple 3D electron microscopy (EM) volumes and instance segmentation masks for neural structures across diverse imaging conditions. These data support the SegNeuron generalist model for cross-volume neuron reconstruction and instance segmentation.
Use Cases
- Train 3D instance segmentation models using the EM volumes and neuron instance masks
- Evaluate model generalization across different EM imaging modalities and tissue types
- Develop automated neuron reconstruction workflows for large-scale connectomics using 3D masks
Strengths
- Includes 3D volumetric data from Electron Microscopy (EM)
- Provides instance-level labels for individual neuron segments
- Aggregates data from multiple EM imaging sources to support generalist model training