Homepage of Alexander Bornik

Description

Passport image

Current position

I am currently a senior researcher at the Ludwig Boltzmann Institute for clinical-forensic imaging.

Liverplanner main project webpage
A surgical resection it the treatment of choice for most patients suffering from liver tumors. The liver contains an arrangement 3 vascular systems, which must be considered in liver surgery in addition to the tumor's location. Thus liver surgery requires careful planning based on a model extracted from a CT scan. The model generation process involves automatic segmentation and generation of 3D models. The outcome of the surgery requires that the3D model represent the anatomy correctly. The developed system allows radiologists to inspect the models in a Hybrid VR setup. Gross segmentation errors can be detected by inspecting the stereoscopically visualized surface models of the liver, the vasculature and the tumor(s) within the CT data context, while small inaccuracies of the model can be detected on a high-resolution 2D tablet PC display. Segmentation refinement is based on the idea, the efforts to correct a 3D model of some segmented object should be small compared to performing a manual or semi-automatic segmentation of hundereds of CT image slices. This is in particular true for recent automatic segmentation algorithms are capable of segmenting most parts of e.g. the liver correctly and robustly. A number of tools to correct the remaining errors in both 3D and 2D have been and are still developed.

See my publication on this page.