* Currently working on the use and construction of Deep Learning methods to optimize image analysis in Digital Pathology * Side activity and previous work on design, optimization and validation of a dynamic computer model for coagulation (blood clotting). * Implementation of an ordinary differential equation (ODE) based model for a part of the coagulation process. * Design of methods to estimate the corresponding model parameters based on limited data, using probabilistic methods, Bayesian statistics and mixed effects models. * Model analysis, e.g. to identify those parameters that most strongly determine the model (simulation) output. * Use the intermediate model analysis results to support the planning of new wet-lab experiments

* PKPD project at the department Molecular Imaging Systems * New algorithms for in vivo quantification of kinetic processes, both on the molecular and the cellular level * Compartment modelling, development of algorithms to estimate kinetic parameters directly from LIST mode PET data. * Design of simulation software to model positron-electron annihilation at the site of the tracer, attenuation and scatter of the resulting photons and subsequent detection of the photons at the detector grid. * Co-authorship on multiple patents related to the described work. * Speech recognition project * Design of new methods that automatically adapt existing speech recognition systems to new speakers, based on a short speech sample. * Design, adaptation and implementation of Hidden Markov Models (HMMs) to achieve this goal. * Integration of produced algorithms in an existing, large Philips software package

* Obtained a PhD in Bayesian Statistics at the department of Neural Networks of the University of Nijmegen (The Netherlands). My work focussed mainly on probabilistic modelling of time series data, in various forms. I developed new methods and algorithms around the topics of * Hidden Markov Models and Mixed Effects Models. * Survival Analysis. * Probabilistic parameter estimation and sampling methods (e.g. Markov Chain Monte Carlo sampling). * Approximate probability distributions for complex models.