AUTOMATIC REGISTRATION OF HISTOLOGICAL SECTIONS

Ignacio Arganda-Carreras

Biocomputing Unit, National Centre of Biotechnology (CSIC). Madrid,
SPAIN Escuela Politicnica Superior. Universidad Autonoma de
Madrid. Madrid, SPAIN

Three-dimensional reconstructions of biological structures from
histological sections involve diverse difficulties. The first problem
in such a process is the need of recovering the original tissue
alignment, which was lost in the sectioning process and distorted due
to the tissue staining and mounting process. Different complex methods
have been proposed to solve the registration problem in 2D, based on
fiduciary markers, elastic deformations of the images, piece-wise
registration, warping, etc. However, their complexity makes these
methods not practical for large images. We will expose in this
informal seminar all the methods applied with this aim by the student
in his search for the most efficient and reliable registration
algorithm.


Ignacio Arganda-Carreras is a PhD student in the PhD program in
"Computer Science and Telecommunications" of the Polytechnic School in
the Universidad Autonoma of Madrid. He is a member of the Biocomputing
Unit in the National Centre of Biotechnology, an interdisciplinary
team of people working on Computational and Structural Biology and
Bioinformatics. He developed all his image registration research in
the group lead by Carlos Ortiz-de-Solsrzano, focused on the study of
the breast cancer progression (http://carlos.lbl.gov).