Automatic mining of fruit fly embryo images
Jia-Yu Pan
Andr&#233; G. R. Balan
Eric P. Xing
Agma Juci Machado Traina
Christos Faloutsos
We present FEMine, an automatic system for image-based gene expression analysis. We perform experiments on the largest publicly available collection of Drosophila ISH (<i>in situ</i> hybridization) images, showing that our FEMine system achieves excellent performance in classification, clustering, and content-based image retrieval. The major innovation of FEMine is the use of automatically discovered latent spatial "themes" of gene expressions, <i>LGEs</i>, in the whole-embryo context, as opposed to patterns in nearly disjoint portions of an embryo proposed in previous methods.
