Participants : Daniel Barbeau [ Correspondant ] , Grégoire Malandain, David Rey.
Isis is a software designed to strictly do volume reconstruction from Histological cross sections. It uses in-house algorithms developped at Asclepios as well as the Insight Toolkit from Kitware.
Histological observations provide direct 2D observations of the anatomical structures and allow for visualisation at a higher resolution than CT or MRI. Isis makes the reconstruction of 3D anatomical volumes from serial slices easier for both clinicians and researchers. Clinicians can perform an automatic registration of the whole stack of images, control the quality of the reconstruction and correct the errors locally or globally. Researchers have access to finer controls to the registration algorithms and could process images before registration. The output of Isis is a volume image and the geometrical transformations between each image.
The work concerning Isis at the end of 2007 and during 2008 covers the definition of its use cases, its functional and technical specifications.