The visualization of data streams plays an important role in diagnosing anomalies in the human body, particularly in Intensive Care Units (ICU). We propose an unconventional paradigm in computer science called Empirical Modelling that is suitable for the combined visualization and exploration of biomedical data streams. Empirical Modelling for Dynamic Visualization (EMDV) provides a learning space for medical education by simulating the ICU patient data streams using web based technology. EMDV benefits from the principles of EM as building models based on observables, dependency and agency, which promotes flexibility and responsiveness in visualizing biomedical data streams. EMDV integrates biomedical signal from different tools into a single monitor. This allows the reduction of cognitive burden in exploring multiple monitors. Open-ended characteristic of EM environment that can enhance the experience of human-machine interaction is founded to be useful to be used as an education technology. The proposed model has been tested with different kinds of mobile devices. The results have clearly shown that the performance of visualization is largely based on the performance of the devices.Index Terms-ICU data streams, visualization, Empirical Modelling, interaction technique, biomedical signal processing, human machine interaction, education technology.
LINTRODUCTIONReal-time observations of streaming data become a critical success factor for many visualization applications [1,4], Many of those applications are using conventional paradigm that encompasses rather fix states of observations. In this research, a well-described definition that conforms to the way human think of the dependencies between observations, has been proposed to create the visualization of data streams. This type of dependencies is conceptually similar to the dependencies between cells in a spreadsheet [2]. Meanwhile, its intertwined relationship could promote a more responsive dynamic visualization that can help human construal process better. With the advances in medical technology, biomedical or clinical data streams are in abundance for exploring and analyzing anomalies or patterns that might lead to early diagnosis. A system that can describe those data streams and provide medical staff with a more flexible user interface for finding anomalies or mathematical patterns could contribute to better healthcare in the future. An example of flexible user interfaces are 978-1-4673-9158-01151$31.00 ©20l5 IEEE