This repository contains code to perform regularised Non-negative Matrix Factorisation (NMF) on two-dimensional time series (movies). It is accompanied by a research paper [1] (open access). In this particular study, we analysed data from Intrinsic Optical Signal (IOS) imaging of the mouse olfactory bulb (OB) during odor stimulation. It demonstrates automatic segmentation of odor-responding neuropils, so-called glomeruli, in the OB.
The release contains the code to perform automatic segmentation using NMF. An IPython notebook demonstrating its use is provided. The notebook can be inspected on NBviewer.
[1] Soelter J, Schumacher J, Spors H, Schmuker M: Automatic segmentation of odor maps in the mouse olfactory bulb using regularized non-negative matrix factorization, NeuroImage, Volume 98, September 2014, Pages 279-288, ISSN 1053-8119, http://dx.doi.org/10.1016/j.neuroimage.2014.04.041.