A new method of detecting organelles on TEM images for better labor and cost efficiency

January 15, 2015

RIKEN CSRS has jointly developed a semi-automatic organelle detection system for transmission electron microscope (TEM) imaging data in collaboration with Tokyo University and Japan Women’s University.

Recent advancements in TEM devices and image processing technologies have enabled acquisition of large-scale datasets, allowing researchers to determine the number and distribution of subcellular ultrastructures at both the cellular and tissue level. At the same time, these advancements have created a new issue: higher costs associated with researchers needing to visually inspect large volumes of image data.

To overcome this issue, the researchers combined the jointly developed CARTA (Clustering-Aided Rapid Training Agent) program with semi-automatic detection procedures to highlight and enlarge regions of interest on images. This method has proved effective for accurately detecting multiple types of organelles such as mitochondria, chloroplasts and Golgi stacks, while reducing the labor required to identify items of interest.

This new system is expected to greatly help accelerate data analysis of TEM imaging data in various research fields while saving manpower and costs.

Original article

Scientific Reports doi: 10.1038/srep07794

T. Higaki, N. Kutsuna, K. Akita, M. Sato, F. Sawaki, M. Kobayashi, N. Nagata, K. Toyooka, S. Hasezawa,

"Semi-automatic organelle detection on transmission electroln microscopic images".

Contact

Kiminori Toyooka

Senior Research Scientist

Gene Discovery Research Group