Research interests: interactive visualisation, virtual reality, user-oriented design, biomedical systems
Exploration of combined multi-modal interaction of virtual reality and desktop systems
Today's interactive exploration environments may vary from non-immersive desktop representations on a conventional PC or a PDA, to fully-immersive CAVE-like virtual reality (VR) environments and augmented reality systems. Desktop and VR systems are two viable alternatives, which allow users to perform manipulations with and navigations through visualised datasets. However, none of them is able to provide optimal means for both visualisation and interaction. It became clearer to us in various projects and experiments, where we focused on the image-based exploration of medical data. Therefore, we decided to concentrate our efforts on the development of a multi-modal visualisation framework, which supports input devices and display configurations of both desktop and VR systems. This framework will allow end-users to alternate desktop and virtual realities while performing interactive steering tasks. Two design concepts of simultaneous and sequential support of realities are being investigated. As a first step, we plan to use the developed framework as an experimental set-up to investigate task-dependency of people's preferences for visual representations, navigation and manipulation methods provided by desktop, VR and mixed reality (desktop and VR) systems. User-steering of semiautomated segmentation of medical images is of the specific interest to us at this moment. Our ultimate goal is to develop a multi-modal visualisation environment capable of adaptive alternating of desktop and virtual realities based on task requirements and subjective users' preferences.
This research is sponsored by NWO/VIEW project "A Multi-modal Visualisation Environment for Interactive Analysis of Medical Data" (643.100.602).
For more information, check this poster or visit our project website.
A Multi-display Environment for the Critical Care Unit
This project is aimed at developing an integrated multi-display solution for the trauma room. The assessment of
the patient's condition requires fast analysis of various data (monitoring data, X-rays, CT scans, ultrasonic data,
results of laboratory test etc.). To get access to this data, members of the Critical Care Unit have to alternate between
different equipment and PC stations not always located in the same room. The goal of this project is to reduce this
overhead by providing clinicians with a possibility to display all required information simultaneously on a large
multi-display system installed directly in the trauma room. The contextual inquiry approach has been chosen for conducting
requirements analysis. We focus on the context of use, technological and functional requirements, users’ skills and knowledge.
This project is conducted in collaboration with the Amsterdam Medical Center and SARA Computing and Networking Services. For more information, check these slides.
Investigation of the practical deployment of Virtual Reality systems in real-life environments
We have combined qualitative methods for the usability evaluation of VR systems and applied them to the development of a prototype of a VR system for medical diagnosis and planning for vascular disorders, the Virtual Radiology Explorer (VRE). The heuristic evaluation and cognitive walkthroughs were applied to assess usability of the VRE prototype and to improve the usability of future versions of the VRE. We also conducted a small exploratory study as a first step to investigate the daily working context of two focus user groups - interventional radiologists and vascular surgeons - to make sure that the VRE is developed in accordance with their real life demands. The whole trajectory of tasks related to diagnosis and treatment planning has been observed in a manner resembling contextual inquiry.
Adaptive interaction in problem-solving environments
A problem-solving environment (PSE) provides users with a set of tools, which may include both software and hardware, for building a specific framework to solve a target class of problems. A PSE should be built in such a way that a user (scientist) might exploit underlying technologies without a specialized knowledge. Unfortunately, in practice the situation is far from ideal. The current research within the PSE community remains focused on high-performance algorithms, security issues and reasonable resource scheduling schemes. As for the interface/interaction adaptation principles, they have not been applied sufficiently in any existing PSE. Therefore, we decided to address this research concern and investigate how adaptive interaction can be organised within a PSE. A distributed PSE is of the most interest to us due to the fact that it permits users to utilize various types of VR systems (e.g., non-immersive desktop, immersive VR, augmented and/or mixed reality) within the same environment.
Design and development of adaptive interfaces based on a user model
Adaptive interface is aimed at helping in the user's work with a system environment by providing a user with an interface that fits him/her the best. Investigation of the influence of personal user characteristics on the process of human-computer interaction based on empirical studies led us to the development of a new methodology of how to control the process of the interface adaptation. We base our approach on two models: 1) an interaction model that describes possible interaction/interface adjustments within a system environment; and 2) a user model, defining a set of human factors guiding this adaptation. We investigated through the set of user studies the influence of selected factors, cognitive and psychological, on people's preferences for different types of graphical user interfaces. The Knowledge Engineer's Workbench (KEW) was chosen for performing the empirical part of the research. To assist the process of user modeling, we developed an instrumental complex. The Intelligent System for User Modeling was used both as a testing environment and as a knowledge acquisition and structuring tool for building the KEW's knowledge base. Based on the experimental results and its statistical evaluation, we formulated hypotheses about the impact of each human factor on the process of the interface adaptation. Later these hypotheses were validated using Monte Carlo simulation method.