Unsupervised cell identification on multidimensional X-ray fluorescence datasets

ABSTRACT

X-ray fluorescence microscopy is a powerful technique to map and quantify trace element distributions in biological specimens. It is perfectly placed to map nanoparticles and nanovectors within cells, at high spatial resolution. Advances in instrumentation, such as faster detectors, better optics, and improved data acquisition strategies are fundamentally changing the way experiments can be carried out, giving us the ability to more completely interrogate samples, at higher spatial resolution, higher throughput and better sensitivity. Yet one thing is still missing: the next generation of data analysis and visualization tools for multidimensional microscopy that can interpret data, identify and classify objects within datasets, visualize trends across datasets and instruments, and ultimately enable researchers to reason with abstraction of data instead of just with images.

References

1. Arteta, C., Lempitsky, V., Noble, J. A., and Zisserman, A. 2012. Learning to detect cells using non-overlapping extremal regions. In Proceedings of the 15th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI12), Springer-Verlag, Berlin, 348--356.
2. Bergeest, J.-P., and Rohr, K. 2012. Efficient globally optimal segmentation of cells in fluorescence microscopy images using level sets and convex energy functionals. Medical Image Analysis 16, 7 (Oct.), 1436--1444.