The histological examination of adjacent tissue sections prepared with different stains or 
biomarkers can provide a valuable amount of information to help understanding of the 
physical or functional properties of tissue at microscopic level. Traditionally 
pathologists and researchers view glass microscopy slides serially with a microscope 
and combine the information mentally to derive an opinion or diagnosis. Although this 
works for global assessment of tissue sections, detailed assessment and measurement 
requires more detailed side-by-side comparison. However, due to the nature of the 
slide preparation and the fact that different stains characterise different substances, 
the tissue sections do not have the same morphology, appearance, or spatial alignment, 
making it a non-trivial task to even find the same region on adjacent slides. The 
introduction of digital pathology (by scanning tissue sections and digitising them 
into histology images) allows the development of automatic computer-aided registration 
algorithms to assist pathologists and researchers quantitatively analysing the spatial 
co-occurrence of structural and functional elements  in different modalities. The main 
challenges for registering histology images with different stains are: 1) dissimilar 
appearances; 2) non-rigid distortions; 3) the large size of digital microscopy images. 
We address these challenges with 1) A multi-resolution block matching based non-rigid 
registration scheme, 2) 2D unsupervised content classification method.  This work is 
funded by Wellcome Trust and the EPSRC, as part of the £11 million project of "fifty 
activate years after fifty".