To study cell function in various contexts, scientists assess interactions between ultrastructural cellular compartments and organelles using volume electron microscopy. To date, scientists had to spend hundreds of hours to manually outline the shape of organelles in just a handful of datasets. Consequently, the manual image annotation represents an important bottleneck for the analysis of 3D electron microscopy images. To overcome this bottleneck, automated image segmentation and analysis is going to be key.
Shortly after the launch of LMtrace®, an online neuron reconstruction service for light microscopy data, the team of ariadne.ai has moved to nanometer-scale resolution and complemented its portfolio by creating the first online cell segmentation workflow for electron microscopy data called 3dEMtrace.
With 3dEMtrace one can segment complex cell organelles, such as endoplasmic reticulum, and even down to the scale of vesicles. Hence, mitochondria-ER contact sites and number of synapses become quantifiable.
“I am very proud of our team of talented biologists and computer vision experts,” says CEO Dr. Adrian Wanner. “Releasing a second high-throughput image analysis pipeline in less than half a year, required tremendous efforts.” Wanner is convinced that 3dEMtrace will unravel important information that was left unanalyzed in the past. 3dEMtrace is the first online cell segmentation service for electron microscopy datasets. It allows insight into a previously hidden part of information and catalyzes the generation of new information by orders of magnitude.
Users can simply upload datasets, select the features to be segmented, and they will receive quality-checked cell segmentation directly to their mailboxes for download.
Just upload your dataset at https://ariadne.ai/3demtrace, and we take it from there!
ariadne-service gmbh is a Swiss company dedicated to the annotation and segmentation of biological tissue and ultrastructures in 3D light and electron microscopy stacks.
We combine human and artificial intelligence to accelerate and facilitate innovation in life science. Our workflows integrate state-of-the-art deep learning algorithms, manual image annotation, and segmentation proofreading by human experts.