MCBMIIA 2015: Abstracts of Invited Talks

摘要

The Radon transform is the mathematical foundation for image reconstruction from projections. Mathematics provides non-invasive methods to measure human body and accurate algorithms to construct slice images of human body. The Radon transform was originally introduced in integral geometry in 1917. Although the transform had initially no relationship to image analysis, the connection between the Radon transform and the Fourier transform allows now us to design accurate and stable algorithms for image reconstruction from projections. Mathematics has, as a result, played a key role diagnosis in modern medicine. The image registration and higher-level segmentation of volumetric and spatiotemporal images are indispensable techniques for pre-processing of computer-aided diagnosis. In early pattern recognition, deformable shape matching was introduced as a two-dimensional extension of the dynamic warping of signals to character recognition. Three-dimensional registration problems are efficiently solved on GPU using modern numerical methods. Recently, there are tremendous progress in image acquisition techniques in multiple scales and multiple modalities. The large amounts of these image data are beyond hand craft works by experts and technicians. The employment of computer vision techniques that interpret spatiotemporal information from correction of low-dimensional information is a key solution to deal with such data in biology. The cell biologists realise that supports by automatic image processing systems allow them to discover new phenomena in cells, understand functionalities in the cell and behaviour of living cells. These are extension of stere-ology to dynamic microscopic images and videos. To totally care human, methodologies on computer vision that allow scientists to continually interpret biological functionalities from cells to organs are desired. The aim of this workshop is to prepare a discussion forum for researchers in computer vision and in biomedical imaging and image analysis to exchange new ideas and problems and to open new application fields in both fields. The forum will derive a bridge on these gaps from the viewpoints of mathematical and computational aspects in computer vision.