Auto-segmentation method


The auto-segmentation method combines semantic segmentation and instance segmentation to produce separate organelle masks. In the soft X-ray tomograms of pancreatic β-cells, we have generated whole-cell masks, mitochondria masks and nucleus masks via semantic segmentation, and insulin vesicle instance masks via instance segmentation.


For more information, please visit SaliLab-SH/Cell-Segmentation



Post-processing tool


The post-processing tool aims to identify and separate individual organelles from semantic organelle masks. It integrates information including intensity of the raw tomogram, semantic segmentation mask of the organelle, and prior knowledge about the organelle morphology. In the soft X-ray tomograms of pancreatic β-cells, we have identified and separated individual organelles including insulin vesicles and mitochondria.


For more information, please visit SaliLab-SH/post_processing_tool



Systematic analysis pipeline


The systematic analysis pipeline characterizes maturation status, localization and distribution of organelles, as well as distance and interactions between organelles. In the soft X-ray tomograms of pancreatic β-cells, we have analyzed the structural variances of nucleus, insulin vesicle, mitochondrion and the whole-cell under different treatment conditions at different time points.


For more information, please visit SaliLab-SH/Cell-Segmentation