We also demonstrate that partial volume is not perfectly modeled, even though methods that account for mixture classes outperform methods that only consider pure Gaussian classes. Finally, we show that simulated data results can also be extended to real data.Index Terms-Brain tissue models, hidden Markov random fields models, magnetic resonance imaging, partial volume, statistical classification, validation study.

This paper introduces two important issues of image registration. At first we want to recall the very general definition of mutual information that allows the choice of various feature spaces to perform image registration. Second we discuss the problem of finding the global maximum in an arbitrary feature space. We used a very general parallel, distributed memory, genetic optimization which turned out to be very robust. We restrict the examples to the context of multi-modal medical image registration but we want to point out that the approach is very general and therefore applicable to a wide range of other applications. The registration algorithm was analysed on a LINUX cluster.

We propose an information theoretic model that unifies a wide range of existing information theoretic signal processing algorithms in a compact mathematical framework. It is mainly based on stochastic processes, Markov chains and error probabilities. The proposed framework will allow us to discuss revealing analogies and differences between several well-known algorithms and to propose interesting extensions resulting directly from our formalism. We will then describe how the theory can be applied to the rapidly emerging field of multi-modal signal processing: we will show how our framework can be efficiently used for multi-modal medical image processing and for joint analysis of multi-media sequences (audio and video). r

Abstract. We developed a software tool for pre-operative simulation and planning, and intra-operative guidance, of minimally invasive tumor ablation, including radiofrequency-, laser-and cryo-therapy. This tool provides a pre-and intra-operative optimization of the treatment plan, in order to avoid dangerous probe trajectories, undertreatment of the tumor, and excessive ablation of healthy tissues. The simulation is performed within a virtual operating-room consisting in essence of the patient's segmented anatomy from pre-or intraoperatively acquired MR scans. Virtual probes can be placed into this scene and at the formation of ablated tissue at their tips can be simulated. To verify the simulated treatment plans, we introduced an objective quality measure which also enables a semi-automated optimal probe placement. To show the use and to underline the importance of our tool, we investigated a cryo-therapy case which did not succeed. We show that our software would have predicted the failure of the chosen treatment plan and how it could have increased the efficacy of the procedure.