Biomedical Engineering Reference
In-Depth Information
Fig. 5.1
Multimodal and multiscalar requirements in the visualization of biomedical data
to match and compare the data according to criteria defined by the scientist using the
system.
Visualization is the primary channel through which biomedical data is communi-
cated. In this work, interaction is a way to control the flow of this channel. It allows
the user to decide in real-time on visualization contents that are to be presented. The
user queries for data through interaction methods, and receives semantic meaningful
results through visualization. Thus, visualization combined with and controlled by
user interaction constitutes in the given context the complete tool for biomedical data
exploration.
The next section discusses the complexity of biomedical data and the require-
ments of a multiscale environment. The third section proposes current lines of work
to design an efficient scientist-centric visualization tool. Current multiscale visual-
ization techniques and the foundation of the Human Computer Interaction (HCI)
field are described in the fourth section. Finally, an overview of strategies and lines
of work on multiscale visualization and interaction with biomedical data is given in
the last section.
5.2 Visualization of Biomedical Data
The presentation of image data has become more challenging due to the increasing
complexity of biomedical datasets. Here it is necessary to integrate all data from dif-
ferent modalities in the same reference system for a specific domain of knowledge
(multimodal requirement). Furthermore, it is also necessary to consider the merg-
ing of several domains across scales (multiscalar requirement). Both of the afore-
mentioned requirements are crucial in the biomedical environment as illustrated in
Fig.
5.1
.
The complexity of data is due to the following factors:
Variety of sources:
Biomedical data can be acquired by a broad range of modal-
ities, e.g. CT, MRI, motion capturing or microscopy. Even sources of the same
modality might not share a common standard. Currently, no standardization of for-
mats in microscopy imaging exists. This leads to a loss of metadata during format
conversion, or problems by organizing images from time-lapse experiments [
1
].
