Speakers: Mattia Natali (Visualization Group, University of Bergen)

Place: Konferanserom B (TM51:KONFB) - VilVite Science Center, Thormølensgate 51

Time: Friday April 26, 2013, from 10.15am to 11.15am

Abstract:
The process of creating terrain and landscape models is important in a variety of computer graphics and visualization applications: from films and computer games, via flight simulators and landscape planning, to scientific visualization and subsurface modelling. Interestingly, the modelling techniques used in this large range of application areas have started to meet in the last years. New trends in geological modelling are approaching the modelling methods that have been developed in computer graphics. I will give an introduction to the process of geological modelling followed by two taxonomies with descriptions and comparisons of selected methods. The first taxonomy of different modelling methods is a data oriented taxonomy, where modelling is categorized into three different scenarios: the data-free, the sparse-data and the dense-data scenario. Then I will show a workflow oriented taxonomy, where modelling is divided into the separate stages necessary for creating a geological model.

Additional material: Flyer

Speakers: Fredrik Orderud (R&D engineer, GE Vingmed Ultrasound - GE Healthcare)

Place: Lille Auditorium, floor 2, Høyteknologisenteret (datablokk)

Time: Friday March 22, 2013, from 10.15am to 11.15am

Abstract:
We present an approach for using state estimation methods for segmentation and real-time tracking of structures in 3D cardiac ultrasound. A Kalman tracking framework is used to compute a least-squares fitting or active-shape surfaces and subdivision surfaces to boundaries in volumetric image data using edge-detection measurements, optionally in combination with speckle-tracking measurements. Typical execution times are 5ms per model per frame on standard computer hardware. Recent edge-detection improvements and biomechanical (FEM) regularization for multi-resolution segmentation will also be covered.

Additional material: Flyer, Fredrik Orderud's webpage

Speakers:
- Astri Johansen Lundervold (Department of Biological and Medical Psychology)
- Arvid Lundervold (Department of Biomedicine)

Place: Lille Auditorium, floor 2, Høyteknologisenteret (datablokk)

Time: Friday March 1, 2013, from 10.15am to 11.15am

Abstract:
Aging is associated with cognitive decline and associated changes in brain structure and function, modulated by genetics, environment, and stochastic processes (arrow of time). Most research on the aging brain has used cross sectional designs, although longitudinal investigations are necessary to uncover more precisely age-related changes in individuals, and how healthy aging differ from early signs of a neurodegenerative disease. Such longitudinal studies have demonstrated large individual differences in subjects between 50 and 80 years of age. Some older individuals preserve both cognitive performance and brain structure and function into high age, others demonstrate early cognitive decline and gross changes in imaging-derived measures, and there is presently an immense interest to identify contributing lifestyle and genetic factors in the aging process. This has also motivated the current study on cognitive aging, brain imaging and genetics, which started in 2003 as a collaboration between researchers at the University of Oslo and the University of Bergen. The subjects in the Bergen cohort have been invited to participate in a baseline and two follow-up studies, and more than 100 individuals from the Bergen area have given their informed consent to take part in a comprehensive neuropsychological examination and structural and functional brain imaging sessions at three separate occasions about three years apart. They have also delivered blood samples for DNA analysis.
In this seminar Astri J. Lundervold will present the study design, the cohort sample, and characteristics of the cognitive data being collected in the study. Arvid Lundervold will present the MR imaging part including the multimodal recordings, image processing and data analysis. Analysis of such heterogeneous cohort data is challenging, both technically and with respect to mutual understanding across the disciplines that are involved in the project. Both presenters will give short examples from the previous and ongoing research and look forward to interesting discussions and input from the audience.

Additional material: Flyer, Astri Lundervold's webpage, Arvid Lundervold's webpage, Astri's slides, Arvid's's slides, Video

Speaker: Ove Daae Lampe (Research scientist, CMR)

Place: Auditorium, VilVite Science Center

Time: Friday January 25, 2013, from 10.15am to 11.15am

Abstract:
Enlighten is Interactive Visual Analysis framework which can read data from multiple sources and provide suitable visualizations using an easy and understandable user interface. Multiple visualizations can be connected/coordinated using interactive filtering. Enlighten is implemented using a mix of OpenCL, C++, R and primarily Python in a modular fashion for extensiblity.
This talk will describe the underlying structure of Enlighten, how to deal with large datasets in Python and how to seamlessly tie together multiple paradigms for computation (Python, OpenCL, OpenGL and R).

Additional material: Flyer, CMR webpage, Slides