Dandan Hu, Pinaki Sarder, Peter Ronhovde, Sharon Bloch, Samuel Achilefu, Zohar Nussinov
We apply a multiresolution community detection algorithm to perform unsupervised segmentation of complex intracellular signals derived using fluorescent dyes. In our earlier work, when applying our method to benchmarks, our algorithm was shown to be one of the best and to be especially suited to the detection of camouflage images. In the current manuscript, we have explored this algorithm in a more complex scenario. The current image processing problem is framed as identifying clusters with respective average fluorescent lifetimes (FLTs) against a background or "solvent" in fluorescence lifetime imaging microscopy (FLIM) images derived using NIR fluorescent dyes. We have identified significant multiresolution structures using replica correlations in these images, where such correlations are manifested by information theoretic overlaps of the independent solutions ("replicas") attained using the proposed algorithm from different starting points. Our method is more efficient than a well-known image segmentation method based on mixture of Gaussian distributions. It offers more than 1.25 times diversity based on Shannon index than the latter method, in selecting clusters with distinct average FLTs in NIR FLIM images.
View original:
http://arxiv.org/abs/1208.4662
No comments:
Post a Comment