Versión de HTML Básico
22
3 Results and discussions
The purpose of this project was to develop and build a system that allows a detailed and flexible display
of a biomedical signal that was previously recorded. The input data are simply numbers and header
information, this is a time series with the numerical value of the signal amplitude at each sampling time [8].
All these numbers contain information that could potentially give the doctor the indication of any pathology
or evidence of physiological phenomena necessary for diagnosis.
The input data which originally resided in a text file type are transferred to a data structure as a whole,
allowing the data to be accessed for future use. The original data set remains unchanged during all
procedures of the system, so there is always the possibility of returning to the source data and to avoid
distortions due to accumulated displacement functions or zoom in / out of the signal plotted on the screen.
Furthermore it is noted that the data have a structure suitable as a list which may facilitate further
processing of the signal beyond viewing. Operations such as digital filtering, amplitude normalization (auto
volume) and data mining to detect patterns and relationships can be applied in future research.
Graphical operations redraw the graph on-screen and for this reason, the speeds of the processor and
video card from the PC which runs the system affect the display of these changes. The system is not
programmed to handle the video memory directly but recalculate the pixel data to graph from functions.
This is not the quickest way to see the changes. For future versions the video memory may be worked
directly to achieve real-time movements provided they are relevant to achieve the diagnostic purpose of the
system.
The sampling frequency with which a biomedical signal was digitized defines the precision with which one
can perform a detailed observation or visual analysis. This precision is called signal resolution. When the
user performs an extreme close-up, the limit of resolution becomes increasingly apparent and could be
interpreted as a limitation of this system when it is actually a limitation of the source data. Future work may
include other improvements to the signal processing itself, keeping in mind things like the Nyquist limitation
for minimum resolution requirements for a particular processing or filtering operation.
Fig. 2.
The system in operation showing the simultaneous display of two biomedical signals.
This software does not require any special database drivers or any licenses in addition to the
programming environment of Microsoft Visual C# [9]. The biomedical signals are stored in the form of text
documents in any folder on a hard drive. The format interpreted in these test signals were in National
Instruments LabView
TM
format (*.LVM). Currently, a text file converter is being programmed to translate
between for Excel
TM
comma-separated *.CSV format and *.LVM.