Abstract:
The thesis contributes to the development of new algorithms, models and applications for medical visualization systems based on the last generation of input and output hardware devices. These components have an important role in biomedical research, medical training, patient diagnosis, surgery planning and treatment monitoring. From the technological point of view the solutions presented in the thesis focus on the gesture based interaction with the rendered 3D models that can be visualized using an autostereoscopic display. This requires a comprehensive analysis of the current software and hardware technologies involved in three-dimensional pseudo-holographic imaging. Also a key aspect is that all the software is browser-based eliminating the need for any specific operating system that in some cases are a barrier for nontechnical users. Due to the limitations imposed by the nature of the involved software technologies a new segmentation algorithm was developed and implemented that allows a faster execution. The software tools where used on publicly available real medical image sets and 3D molecular models in order to validate the results based on a proposed methodology and specific test scenarios for students in medicine and chemistry.
