Multi-modal Interaction in Biomedicine - Ambient Intelligence for Scientific Discovery - page 184

Multi-modal Interaction in Biomedicine

Elena V. Zudilova and Peter M.A. Sloot

University of Amsterdam, The Netherlands
{ elenaz,sloot } @science.uva.nl

1 Introduction

Everybody agrees that user tasks and preferences should play a central role in
the design and development of applications oriented to non-computer experts.
Nevertheless, even biomedical applications are sometimes developed in a rela-
tive vacuum from the real needs of end-users and environments where they are
supposed to be used.
To provide a clinician with an intuitive environment to solve a target class
of problems, a biomedical application has to be built in such a way that a user
can exploit modern technologies without specialized knowledge of underlying
hardware and software [18]. Unfortunately, in reality the situation is different.
Many developers do not take into account the fact that their potential users
are people, who are mostly inexperienced computer users, and as a result they
need intuitive interaction capabilities and a relevant feedback adapted to their
knowledge and skills.
User comfort is very important for the success of any software application
[13]. But very often we forget that usability problems may arise not only from
an 'uncomfortable' graphical user interface (GUI), but also from a projection
modality chosen incorrectly for deploying an interactive environment [16].
Existing projection modalities have not been suciently investigated yet in
respect to usability factors. Meanwhile, the selection of an appropriate projection
modality in accordance with the user's tasks, preferences and personal features
might help in building a motivated environment for biomedical purposes. In this
chapter we summarize our recent findings related to this research and introduce
a new concept of multi-modal interaction based on the combination of virtual
reality (VR) and desktop projection modalities within the same system. For the
case study of the research we used a biomedical application simulating vascular
reconstruction [2, 22].
The rest of the chapter is organized as follows. Section 2 introduces con-
cepts of a multi-modal interaction and projection modalities. Section 3 describes
the biomedical application for vascular reconstruction deployed for two different
projection modalities. Section 4 is devoted to the experiments on user profiling.
Both the methodology, on which the user profiling was based, and the results
are presented here. In section 5 the possibilities of how VR and desktop projec-
tion modalities can be combined are discussed. Finally, conclusions and plans
for future research are presented in section 6.