3-4 hours of lectures per week.
Lecturer
Content
New advanced mathematical, statistical and computational methods for data handling, especially within the discipline of image analysis, have opened up the possibility for signi cant improvement of the quality of the diagnostic information extractable from medical images. In particular, the recent theoretical advances in computational stochastics and scienti c computing (development of MCMC simulation techniques and PC cluster computers) make it possible to analyze highly complex image models and handle huge data sets involved in dynamic imaging in real time. This course will give an introduction to these techniques. Only a basic knowledge of statistics is required. The course consists of two parts of roughly the same extent. The rst part is an introduction to simulation. Traditional techniques such as rejection and importance sampling will be mentioned. Also modern techniques such as MCMC simulation will be studied. In the second part of the course, selected low and high level image models will be studied in detail. In particular, it will be shown how these models can be simulated and used in the analysis of images, including in image segmentation. Lecture notes will be provided.Prerequisites
Biostatistics, Geostatistics or Statistics Alpha.